diff --git a/.gitignore b/.gitignore
index e69de29..408286c 100644
--- a/.gitignore
+++ b/.gitignore
@@ -0,0 +1,7 @@
+work/*
+runs/*
+logs/*
+*log
+.nextflow.log*
+.nextflow*
+results/*
diff --git a/bin/Metafusion_forte.sh b/bin/Metafusion_forte.sh
new file mode 100755
index 0000000..53874b6
--- /dev/null
+++ b/bin/Metafusion_forte.sh
@@ -0,0 +1,159 @@
+#!/bin/bash
+#STEPS
+
+# __author__      = "Alexandria Dymun"
+# __email__       = "pintoa1@mskcc.org"
+# __contributor__ = "Anne Marie Noronha (noronhaa@mskcc.org)"
+# __version__     = "0.0.1"
+# __status__      = "Dev"
+
+
+output_ANC_RT_SG=1
+RT_call_filter=1
+blck_filter=1
+ANC_filter=1
+usage() {
+    echo "Usage: Metafusion_forte.sh --num_tools=<minNumToolsCalled> --genome_fasta <FASTA adds SEQ to fusion>  --recurrent_bedpe <blacklistFusions>  --outdir <outputDirectory> --cff <cffFile> --gene_bed <geneBedFile> --gene_info <geneInfoFile>" 1>&2;
+    exit 1;
+}
+
+# Loop through arguments and process them
+while test $# -gt 0;do
+    case $1 in
+        -n=*|--num_tools=*)
+        num_tools="${1#*=}"
+        shift
+        ;;
+        --outdir)
+        outdir="$2"
+        shift 2
+        ;;
+        --cff)
+        cff="$2"
+        shift 2
+        ;;
+        --gene_bed)
+        gene_bed="$2"
+        shift 2
+        ;;
+        --gene_info)
+        gene_info="$2"
+        shift 2
+        ;;
+        --genome_fasta)
+        genome_fasta="$2"
+        shift 2
+        ;;
+        --recurrent_bedpe)
+        recurrent_bedpe="$2"
+        shift 2
+        ;;
+        *)
+        #OTHER_ARGUMENTS+=("$1")
+        shift # Remove generic argument from processing
+        ;;
+    esac
+done
+
+if [[ ! $cff || ! $gene_info  || ! $gene_bed ]]; then
+    echo "Missing required argument"
+    usage
+fi
+
+
+mkdir $outdir
+
+#Check CFF file format:
+#Remove entries with nonconformming chromosome name
+
+all_gene_bed_chrs=`awk -F '\t' '{print $1}' $gene_bed | sort | uniq | sed 's/chr//g '`
+awk -F " " -v arr="${all_gene_bed_chrs[*]}" 'BEGIN{OFS = "\t"; split(arr,arr1); for(i in arr1) dict[arr1[i]]=""} $1 in dict && $4 in dict' $cff >  $outdir/$(basename $cff).cleaned_chr
+grep -v -f $outdir/$(basename $cff).cleaned_chr $cff > problematic_chromosomes.cff
+cff=$outdir/$(basename $cff).cleaned_chr
+
+#Rename cff
+echo Rename cff
+rename_cff_file_genes.MetaFusion.py $cff $gene_info > $outdir/$(basename $cff).renamed
+cff=$outdir/$(basename $cff).renamed
+
+#Annotate cff
+if [ $genome_fasta ]; then
+    echo Annotate cff, extract sequence surrounding breakpoint
+    reann_cff_fusion.py --cff $cff --gene_bed $gene_bed --ref_fa $genome_fasta > $outdir/$(basename $cff).reann.WITH_SEQ
+else
+    echo Annotate cff, no extraction of sequence surrounding breakpoint
+    reann_cff_fusion.py --cff $cff --gene_bed $gene_bed > $outdir/$(basename $cff).reann.NO_SEQ
+fi
+
+# Assign .cff based on SEQ or NOSEQ
+if [ $genome_fasta ]; then
+    cff=$outdir/$(basename $cff).reann.WITH_SEQ
+    echo cff $cff
+else
+    cff=$outdir/$(basename $cff).reann.NO_SEQ
+    echo cff $cff
+fi
+
+echo Add adjacent exons to cff
+extract_closest_exons.py $cff $gene_bed $genome_fasta  > $outdir/$(basename $cff).exons
+
+# assign cff as ".exons" if --annotate_exons flag was specified
+
+cff=$outdir/$(basename $cff).exons
+
+
+#Merge
+cluster=$outdir/$(basename $cff).cluster
+echo Merge cff by genes and breakpoints
+RUN_cluster_genes_breakpoints.sh $cff $outdir > $cluster
+
+#output ANC_RT_SG file
+if [ $output_ANC_RT_SG -eq 1 ]; then
+    echo output cis-sage.cluster file
+    output_ANC_RT_SG.py $cluster > $outdir/cis-sage.cluster
+fi
+
+#ReadThrough Callerfilter
+if [ $RT_call_filter -eq 1 ]; then
+    echo ReadThrough, callerfilter $num_tools
+    cat $cluster | grep ReadThrough > $outdir/$(basename $cluster).ReadThrough
+    callerfilter_num.py --cluster $cluster  --num_tools $num_tools > $outdir/$(basename $cluster).callerfilter.$num_tools
+    callerfilter_excluded=$(comm -13 <(cut -f 22 $outdir/$(basename $cluster).callerfilter.$num_tools | sort | uniq) <(cut -f 22 $cluster | sort | uniq))
+    grep -v ReadThrough $outdir/$(basename $cluster).callerfilter.$num_tools > $outdir/$(basename $cluster).RT_filter.callerfilter.$num_tools
+    cluster_RT_call=$outdir/$(basename $cluster).RT_filter.callerfilter.$num_tools
+fi
+# Blocklist Filter
+if [ $recurrent_bedpe ]; then
+    echo blocklist filter
+    blocklist_filter_recurrent_breakpoints.sh $cff $cluster_RT_call $outdir $recurrent_bedpe > $outdir/$(basename $cluster).RT_filter.callerfilter.$num_tools.blck_filter
+
+    blocklist_cluster=$outdir/$(basename $cluster_RT_call).BLOCKLIST
+
+    cluster=$outdir/$(basename $cluster).RT_filter.callerfilter.$num_tools.blck_filter
+fi
+# Adjacent Noncoding filter
+if [ $ANC_filter -eq 1 ]; then
+    echo ANC adjacent noncoding filter
+    filter_adjacent_noncoding.py $cluster > $outdir/$(basename $cluster).ANC_filter
+
+    cluster=$outdir/$(basename $cluster).ANC_filter
+fi
+#Rank and generate final.cluster
+echo Rank and generate final.cluster
+rank_cluster_file.py $cluster > $outdir/final.n$num_tools.cluster
+cluster=$outdir/final.n$num_tools.cluster
+### Generate filtered FID file
+#out=`awk -F '\t' '{print $15}' $cluster  | tail -n +2`
+#out2=`awk -F '\t' '{print $22}' $outdir/cis-sage.cluster | tail -n +2`
+#out3=`echo $out $out2`
+#echo ${out3//,/ } > out4
+#out5=`tr ' ' '\n' < out4 | sort | uniq`
+
+#for this in $(echo $out5); do grep $this $cff; done >> $outdir/$(basename $cff).filtered.cff
+
+rm -f filters.txt
+cut -f 22 *.BLOCKLIST | tr "," "\n" | sort | uniq | sed "s/$/\tblocklist/g" > filters.txt
+cut -f 22 *.ANC_filter | tr "," "\n" | sort | uniq | sed "s/$/\tadjacent_noncoding/g" >> filters.txt
+cut -f 22 *.ReadThrough | tr "," "\n" | sort | uniq | sed "s/$/\tread_through/g" >> filters.txt
+echo -en "$callerfilter_excluded" | tr "," "\n" | sort | uniq | sed "s/$/\tcaller_filter/g" >> filters.txt
+
diff --git a/bin/add_annotations_cff.R b/bin/add_annotations_cff.R
new file mode 100755
index 0000000..7ec1489
--- /dev/null
+++ b/bin/add_annotations_cff.R
@@ -0,0 +1,109 @@
+#!/usr/local/bin/Rscript
+# __author__      = "Anne Marie Noronha"
+# __email__       = "noronhaa@mskcc.org"
+# __version__     = "0.0.1"
+
+
+suppressPackageStartupMessages({
+    library(dplyr)
+    library(data.table)
+})
+
+usage <- function() {
+    message("Usage:")
+    message("add_annotations_cff.R --cff-file <file.cff> --agfusion-file <agfusion.tsv> --oncokb-file <oncokb.tsv> --out-prefix <prefix>")
+}
+
+args = commandArgs(TRUE)
+
+if (is.null(args) | length(args)<1) {
+    usage()
+    quit()
+}
+
+#' Parse out options from a string without recourse to optparse
+#'
+#' @param x Long-form argument list like --opt1 val1 --opt2 val2
+#'
+#' @return named list of options and values similar to optparse
+
+parse_args <- function(x){
+    args_list <- unlist(strsplit(x, ' ?--')[[1]])[-1]
+    args_vals <- lapply(args_list, function(x) scan(text=x, what='character', quiet = TRUE))
+    # Ensure the option vectors are length 2 (key/ value) to catch empty ones
+    args_vals <- lapply(args_vals, function(z){ length(z) <- 2; z})
+
+    parsed_args <- structure(lapply(args_vals, function(x) x[2]), names = lapply(args_vals, function(x) gsub('-','_',x[1])))
+    parsed_args[! is.na(parsed_args)]
+}
+
+args_opt <- parse_args(paste(args,collapse=" "))
+
+possible_args = c("cff", "oncokb", "agfusion", "out_prefix")
+if (length(setdiff(names(args_opt),possible_args)) > 0){
+    message("Invalid options")
+    usage()
+    quit()
+}
+
+if (length(setdiff(possible_args,names(args_opt))) > 0) {
+    message("Missing required arguments")
+    usage()
+    quit()
+}
+
+oncokb_file = args_opt$oncokb
+agfusion_file = args_opt$agfusion
+cff_file = args_opt$cff
+out_prefix = args_opt$out_prefix
+
+cff = fread(cff_file)
+final_cff_cols <- c(names(cff))
+agfusion_tab = fread(agfusion_file) %>% select(c(`5'_transcript`,`3'_transcript`,`5'_breakpoint`,`3'_breakpoint`,Fusion_effect))
+final_cff_cols <- c(final_cff_cols,"Fusion_effect")
+if (!is.null(oncokb_file)){
+    oncokb_tab = fread(oncokb_file) %>% select(-Fusion)
+    final_cff_cols = c(final_cff_cols,names(oncokb_tab %>% select(-Tumor_Sample_Barcode)))
+    cff <- merge(
+        cff,
+        oncokb_tab,
+        by.x ="FID",
+        by.y = "Tumor_Sample_Barcode",
+        all.x = T,
+        all.y=F
+    )
+}
+
+cff <- merge(
+    cff,
+    agfusion_tab,
+    by.x = c("gene5_transcript_id","gene3_transcript_id","gene5_breakpoint","gene3_breakpoint"),
+    by.y = c("5'_transcript","3'_transcript","5'_breakpoint","3'_breakpoint"),
+    all.x = T,
+    all.y = T
+)
+
+cff <- as.data.frame(cff)[,c(final_cff_cols)]
+#cff <- cff %>% mutate(!!final_cff_cols[34] := Fusion_effect) %>% select(-c(Fusion_effect))
+
+write.table(
+    cff,
+    paste0(out_prefix, ".unfiltered.cff"),
+    row.names = F,
+    quote = F,
+    sep = "\t",
+    col.names = ! "V1" %in% final_cff_cols
+)
+
+filtered_cff <- cff %>% filter(! (is.na(cluster) | is.null(cluster) | cluster == ""))
+write.table(
+    filtered_cff,
+    paste0(out_prefix, ".final.cff"),
+    row.names = F,
+    append = F,
+    quote = F,
+    sep = "\t",
+    col.names = ! "V1" %in% final_cff_cols
+)
+
+
diff --git a/bin/add_flags_and_cluster_information.R b/bin/add_flags_and_cluster_information.R
new file mode 100755
index 0000000..b74ff92
--- /dev/null
+++ b/bin/add_flags_and_cluster_information.R
@@ -0,0 +1,199 @@
+#!/usr/local/bin/Rscript
+# __author__      = "Alexandria Dymun"
+# __email__       = "pintoa1@mskcc.org"
+# __contributor__ = "Anne Marie Noronha (noronhaa@mskcc.org)"
+# __version__     = "0.0.1"
+# __status__      = "Dev"
+
+
+
+library(dplyr)
+library(data.table)
+    args <- commandArgs(TRUE)
+    if (length(args) != 6) {
+        stop(
+            "6 arguments are required as input in the following order: unfiltered_cff cluster_file cis_sage_file problematic_chromosomes_file filters_table sample_name"
+        )
+    }
+
+    unfiltered_cff <- fread(args[1],data.table = F)
+    header_cff <-
+        c(
+            "gene5_chr",
+            "gene5_breakpoint",
+            "gene5_strand",
+            "gene3_chr",
+            "gene3_breakpoint",
+            "gene3_strand",
+            "library",
+            "sample",
+            "T_N",
+            "disease",
+            "tool",
+            "max_split_cnt",
+            "max_span_cnt",
+            "gene5_renamed_symbol",
+            "gene5_tool_annotation",
+            "gene3_renamed_symbol",
+            "gene3_tool_annotation",
+            "FusionType",
+            "reann_gene5_symbol",
+            "reann_gene5_region",
+            "reann_gene3_symbol",
+            "reann_gene3_region",
+            "reann_gene5_on_bndry",
+            "reann_gene5_close_to_bndry",
+            "reann_gene3_on_bndry",
+            "reann_gene3_close_to_bndry",
+            "score",
+            "coding_id_distance",
+            "gene_interval_distance",
+            "dnasupp",
+            "FID",
+            "gene5_seq",
+            "gene3_seq",
+            "is_inframe",
+            "closest_exon5",
+            "closest_exon3",
+            "captured_reads",
+            "gene5_transcript_id",
+            "gene3_transcript_id"
+        )
+    colnames(unfiltered_cff) <- header_cff
+    cluster <- fread(args[2],data.table = F)
+    header_cluster <-
+    c(
+        "gene5_renamed_symbol",
+        "gene3_renamed_symbol",
+        "gene5_chr",
+        "gene5_breakpoint",
+        "gene3_chr",
+        "gene3_breakpoint",
+        "max_split_cnt",
+        "max_span_cnt",
+        "T_N",
+        "disease",
+        "tool",
+        "FusionType",
+        "sample",
+        "cancer_db_hits",
+        "FID"
+    )
+    colnames(cluster) <- header_cluster
+    cluster <- cluster %>%
+        mutate(CID=as.character(row_number())) %>%
+        select(c(FID,CID)) %>%
+        tidyr::separate_rows(FID,sep=",")
+
+
+    sample_name <- args[6]
+
+    cis_sage <- tryCatch({fread(args[3],data.table = F)},warning = function(cond){return( NULL)})
+    if(!is.null(cis_sage)){
+        header_cis <-
+        c(
+            "test",
+            "gene5_renamed_symbol",
+            "gene3_renamed_symbol",
+            "max_split_cnt",
+            "max_span_cnt",
+            "T_N",
+            "disease",
+            "tool",
+            "FusionType",
+            "reann_gene5_on_bndry",
+            "reann_gene5_close_to_bndry",
+            "reann_gene3_on_bndry",
+            "reann_gene3_close_to_bndry",
+            "is_inframe",
+            "sample",
+            "gene5_chr",
+            "gene5_breakpoint",
+            "gene3_chr",
+            "gene3_breakpoint",
+            "closest_exons5",
+            "closest_exons3",
+            "FID"
+        )
+
+        colnames(cis_sage) <- header_cis
+
+        cis_sage <- cis_sage %>%
+            mutate(CID=paste0("cis_sage_",row_number())) %>%
+            select(c(FID,CID)) %>%
+            tidyr::separate_rows(FID,sep=",")
+
+        df_cluster <- cluster %>%
+            bind_rows(cis_sage) %>%
+            group_by(FID) %>%
+            summarise(cluster=paste(CID,collapse=';'))
+
+        unfiltered_cff <- merge(unfiltered_cff, df_cluster, by = "FID", all.x = T)
+    }
+
+    unfiltered_cff$Metafusion_flag <- apply(unfiltered_cff, 1, function(row) {
+    if ((is.na(row["reann_gene5_symbol"]) ||
+        is.na(row["reann_gene3_symbol"]))  & (row["gene5_renamed_symbol"] != "." ||
+        row["gene3_renamed_symbol"]  != ".") ) {
+        return("Gene_or_loc_not_in_bed")
+    } else if (row["gene3_renamed_symbol"] != row["reann_gene3_symbol"] ||
+                row["gene5_renamed_symbol"] != row["reann_gene5_symbol"]) {
+        return("Annot_gene_different_from_renamed")
+    } else{
+        return(NA)
+    }
+    })
+
+    weird_chromosomes <- tryCatch({fread(args[4],data.table = F)},warning = function(cond){return( NULL)})
+    if(!is.null(weird_chromosomes)){
+        colnames(weird_chromosomes) <-
+        c(
+            "gene5_chr",
+            "gene5_breakpoint",
+            "gene5_strand",
+            "gene3_chr",
+            "gene3_breakpoint",
+            "gene3_strand",
+            "library",
+            "sample",
+            "T_N",
+            "disease",
+            "tool",
+            "max_split_cnt",
+            "max_span_cnt",
+            "gene5_renamed_symbol",
+            "gene5_tool_annotation",
+            "gene3_renamed_symbol",
+            "gene3_tool_annotation"
+        )
+        weird_chromosomes[, colnames(unfiltered_cff)[!colnames(unfiltered_cff) %in% colnames(weird_chromosomes)]] <- NA
+        weird_chromosomes$Metafusion_flag <- "Chromosome_not_in_bed"
+        unfiltered_cff <- rbind(unfiltered_cff,weird_chromosomes)
+    }
+
+
+    filters <- tryCatch(
+        {
+            fread(
+                args[5],
+                data.table = F,
+                col.names=c("FID","tmpflag")
+            ) %>% group_by(FID) %>% summarise(tmpflag=paste(tmpflag,collapse=''))
+        },
+        warning = function(cond){return( NULL)}
+    )
+    if (!is.null(filters)){
+        unfiltered_cff <- merge(unfiltered_cff,filters, by="FID", all.x = T, all.y = F) %>%
+            mutate(Metafusion_flag=ifelse(is.null(Metafusion_flag) | is.na(Metafusion_flag) | Metafusion_flag=="", tmpflag, paste(Metafusion_flag,tmpflag,sep=","))) %>%
+            select(-c(tmpflag))
+    }
+
+    write.table(
+    unfiltered_cff[,c(header_cff,"Metafusion_flag","cluster")],
+    paste0(sample_name, "_metafusion_cluster.unfiltered.cff"),
+    row.names = F,
+    append = F,
+    quote = F,
+    sep = "\t"
+    )
+
diff --git a/bin/final_generate_v75_gene_bed.R b/bin/final_generate_v75_gene_bed.R
new file mode 100644
index 0000000..57ecbe9
--- /dev/null
+++ b/bin/final_generate_v75_gene_bed.R
@@ -0,0 +1,113 @@
+
+# __author__      = "Alexandria Dymun"
+# __email__       = "pintoa1@mskcc.org"
+# __contributor__ = "Anne Marie Noronha (noronhaa@mskcc.org)"
+# __version__     = "0.0.1"
+# __status__      = "Dev"
+
+
+
+library(dplyr)
+library(data.table)
+library(stringr)
+gtf <- rtracklayer::import('/work/taylorlab/cmopipeline/mskcc-igenomes/igenomes/Homo_sapiens/Ensembl/GRCh37/Annotation/Genes/genes.gtf')
+gtf_df <- as.data.frame(gtf)
+
+
+# Utilized gtf from igenomes for FORTE This corresponds to GRCh37 ensembl 75
+# Add introns to gtf, convert to gff3
+# bsub -R "rusage[mem=64]" -o add_introns_agat_%J.out singularity exec -B /juno/ \\
+# -B /tmp -B /scratch/ docker://quay.io/biocontainers/agat:0.8.0--pl5262hdfd78af_0  \\
+# /bin/bash -c "agat_sp_add_introns.pl -g /juno/work/taylorlab/cmopipeline/mskcc-igenomes/igenomes/Homo_sapiens/Ensembl/GRCh37/Annotation/Genes/genes.gtf\\
+# -o genes.INTRONS.gff3"
+# gff2bed < genes.INTRONS.gff3 > genes.INTRONS.agat.bed
+
+
+total.introns.bed <- fread(file="/work/ccs/pintoa1/metafusion_refs/meta_fusion_bed_generation/genes.INTRONS.agat.bed", header = FALSE, stringsAsFactors = F, sep="\t", na.strings = "",data.table = F)
+colnames(total.introns.bed) <- c("chr","start","end","gene_id","tmp","strand","gene_biotype","type","V9","description")
+total.introns.bed$transcript_id <- gsub("\\;.*","",str_split_fixed(total.introns.bed$description,"transcript_id=",n=2)[,2])
+total.introns.bed$gene_name <-gsub("\\;.*","",str_split_fixed(total.introns.bed$description,"gene_name=",n=2)[,2])
+
+transcript_ids <- unique(total.introns.bed$transcript_id)
+file.to_write <- "/work/ccs/pintoa1/metafusion_refs/meta_fusion_bed_generation/cleaned_metafusion_v75_gene.bed"
+
+if(file.exists(file.to_write) ) {file.remove(file.to_write)}
+
+#START CLOCK: THE INDEXING TAKES A LONG TIME, LIKE 5 HOURS
+ptm <- proc.time()
+
+# Index each transcript feature, incrementing when an intron is passed
+## metafusion expects exon count 0 to (N(exons)-1)
+## Forward strand: Exon 0 == Exon 1
+### Reverse strand: Exon 0 == LAST EXON IN TRANSCRIPT
+for (id in transcript_ids){
+    transcript <- total.introns.bed[total.introns.bed$transcript_id == id,]
+    # Remove exons if coding gene, since "exon" and "CDS" are duplicates of one another
+    if ("CDS" %in% transcript$type){
+        transcript <- transcript[!transcript$type == "exon",]
+        }
+    # Order features by increasing bp
+    transcript <- transcript[order(transcript$start, decreasing = FALSE),]
+    # Index features
+    idx <- 0
+    for (i in 1:nrow(transcript)){
+        transcript$idx [i]<- idx
+        if (transcript$type[i] == "intron"){
+            idx <- idx + 1
+        }
+    }
+    # REFORMAT TRANSCRIPT
+    #Change strand info (+ --> f, - --> r)
+    if (unique(transcript$strand) == "+"){
+        transcript$strand <- 'f'
+    } else if  (unique(transcript$strand) == "-"){
+        transcript$strand <- 'r'
+    } else {
+        errorCondition("Strand info for this transcript is inconsistent")
+    }
+    #Add "chr" prefix to chromosomes
+    transcript$chr <- sapply("chr", paste0,  transcript$chr)
+    #Change CDS --> cds ### IF A TRANSCRIPT LACKS "CDS" THIS LINE WILL DO NOTHING, Changing exon values to UTRs later
+    if ("CDS" %in% unique(transcript$type)){transcript[transcript$type == "CDS",]$type <- "cds"}
+    ## DETERMING UTR3 and UTR5
+    ### INSTEAD OF START AND STOP, USE CDS LOCATIONS AND STRAND INFORMATION.....
+    if ("UTR" %in% unique(transcript$type)){
+        if( unique(transcript$strand) == "f"){
+            #Forward strand
+            start_coding <- min(transcript[transcript$type == "cds","start"])
+            stop_coding <-  max(transcript[transcript$type == "cds","end"])
+            transcript$type[transcript$end <= start_coding &  transcript$type == "UTR"] <- "utr5"
+            transcript$type[transcript$start >= stop_coding & transcript$type == "UTR"] <- "utr3"
+        }else {
+            start_coding <- max(transcript[transcript$type == "cds","end"])
+            stop_coding <- min(transcript[transcript$type == "cds","start"])
+            transcript$type[transcript$end <= start_coding &  transcript$type == "UTR"] <- "utr3"
+            transcript$type[transcript$start >= stop_coding & transcript$type == "UTR"] <- "utr5"
+        }
+    }
+    transcript <- transcript[,c("chr", "start", "end", "transcript_id", "type", "idx", "strand", "gene_name", "gene_id" )]
+    write.table(transcript, file.to_write, append=TRUE, sep="\t", quote=F,  row.names=F, col.names=F)
+}
+
+time <- proc.time() - ptm
+time
+#
+# user    system   elapsed
+# 16657.116    32.227 16741.382
+
+
+new.bed <- fread(file.to_write,data.table = F)
+colnames(new.bed) <- c("chr","start","end","transcript_id","type","idx","strand","gene_name","gene_id")
+
+#### Any exon that remains after teh cds change, is likely and untranslated region. change below
+
+# Basically, subfeatures which are "exon" need to be changed (i.e. exon --> utr3/utr5)
+#Forward strand
+new.bed$type[new.bed$strand == "f" &  new.bed$type == "exon" ] <- "utr5"
+#Reverse strand
+new.bed$type[new.bed$strand == "r" &  new.bed$type == "exon"]<- "utr3"
+
+expected_types <- c("cds","intron","utr3","utr5")
+new.bed.ready <- new.bed[new.bed$type %in% c(expected_types),]
+
+write.table(new.bed.ready, "/work/ccs/pintoa1/metafusion_refs/meta_fusion_bed_generation/v75_gene.bed",  sep="\t", quote=F,  row.names=F, col.names=F)
diff --git a/bin/make_cff_from_forte.R b/bin/make_cff_from_forte.R
new file mode 100755
index 0000000..f000bac
--- /dev/null
+++ b/bin/make_cff_from_forte.R
@@ -0,0 +1,176 @@
+#!/usr/local/bin/Rscript
+
+# __author__      = "Alexandria Dymun"
+# __email__       = "pintoa1@mskcc.org"
+# __contributor__ = "Anne Marie Noronha (noronhaa@mskcc.org)"
+# __version__     = "0.0.1"
+# __status__      = "Dev"
+
+
+library(dplyr)
+library(data.table)
+library(stringr)
+cff_format <-
+    c(
+    "chr1",
+    "pos1",
+    "strand1",
+    "chr2",
+    "pos2",
+    "strand2",
+    "library",
+    "sample_name",
+    "sample_type",
+    "disease",
+    "tool",
+    'split_cnt',
+    "span_cnt",
+    "t_gene1",
+    "t_area1",
+    "t_gene2",
+    "t_area2"
+    )
+
+
+opt <- commandArgs(TRUE)
+
+cff_format_df <-
+    setNames(data.frame(matrix(
+    ncol = length(cff_format), nrow = 0
+    )), cff_format)
+
+make_arriba <- function(sample_file) {
+    df <- as.data.frame(matrix(ncol = 0, nrow = nrow(sample_file)))
+
+    df$t_gene1 <- sample_file$gene_id1
+    df$t_gene2 <- sample_file$gene_id2
+    df$chr1 <- str_split_fixed(sample_file$breakpoint1, ":", 2)[, 1]
+    df$pos1 <- str_split_fixed(sample_file$breakpoint1, ":", 2)[, 2]
+    df$chr2 <-  str_split_fixed(sample_file$breakpoint2, ":", 2)[, 1]
+    df$pos2 <- str_split_fixed(sample_file$breakpoint2, ":", 2)[, 2]
+    df$strand1 <-
+        str_split_fixed(sample_file$`strand1(gene/fusion)`, "/", 2)[, 1]
+    df$strand2 <-
+        str_split_fixed(sample_file$`strand2(gene/fusion)`, "/", 2)[, 1]
+    df$tool <- "arriba"
+    df$split_cnt <-
+        ifelse(!is.na(sample_file$split_reads1),
+            sample_file$split_reads1,
+            -1)
+    df$span_cnt <-
+        ifelse(!is.na(sample_file$discordant_mates),
+            sample_file$discordant_mates,
+            -1)
+    df$t_area1 <- sample_file$site1
+    df$t_area2 <- sample_file$site2
+
+    return(df)
+}
+
+make_fusioncatcher <- function(sample_file) {
+    df <- as.data.frame(matrix(ncol = 0, nrow = nrow(sample_file)))
+
+    df$t_gene1 <- sample_file[, "Gene_1_id(5end_fusion_partner)"]
+    df$t_gene2 <- sample_file[, "Gene_2_id(3end_fusion_partner)"]
+    df$chr1 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_1(5end_fusion_partner)"], ":", 3)[, 1]
+    df$pos1 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_1(5end_fusion_partner)"], ":", 3)[, 2]
+    df$chr2 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_2(3end_fusion_partner)"], ":", 3)[, 1]
+    df$pos2 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_2(3end_fusion_partner)"], ":", 3)[, 2]
+    df$strand1 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_1(5end_fusion_partner)"], ":", 3)[, 3]
+    df$strand2 <-
+        str_split_fixed(sample_file[, "Fusion_point_for_gene_2(3end_fusion_partner)"], ":", 3)[, 3]
+    df$tool <- "fusioncatcher"
+    df$split_cnt <-
+        ifelse(
+            !is.na(sample_file$Spanning_unique_reads),
+            sample_file$Spanning_unique_reads,
+            -1
+        )
+    df$span_cnt <-
+        ifelse(!is.na(sample_file$Spanning_pairs),
+            sample_file$Spanning_pairs,
+            -1)
+    df$t_area1 <- sample_file$Predicted_effect
+    df$t_area2 <- sample_file$Predicted_effect
+
+
+    return(df)
+
+}
+
+make_starfusion <- function(sample_file) {
+    df <- as.data.frame(matrix(ncol = 0, nrow = nrow(sample_file)))
+    df$t_gene1 <- str_split_fixed(sample_file$LeftGene, "\\^", 2)[, 2]
+    df$t_gene2 <- str_split_fixed(sample_file$RightGene, "\\^", 2)[, 2]
+    df$chr1 <-
+        str_replace(str_split_fixed(sample_file$LeftBreakpoint, ":", 3)[, 1],
+                "chr",
+                "")
+    df$pos1 <- str_split_fixed(sample_file$LeftBreakpoint, ":", 3)[, 2]
+    df$chr2 <-
+        str_replace(str_split_fixed(sample_file$RightBreakpoint, ":", 3)[, 1],
+                "chr",
+                "")
+    df$pos2 <-  str_split_fixed(sample_file$RightBreakpoint, ":", 3)[, 2]
+    df$strand1 <-
+        str_split_fixed(sample_file$LeftBreakpoint, ":", 3)[, 3]
+    df$strand2 <-
+        str_split_fixed(sample_file$RightBreakpoint, ":", 3)[, 3]
+    df$tool <- "starfusion"
+    df$split_cnt <-
+        ifelse(!is.na(sample_file$JunctionReadCount),
+            sample_file$JunctionReadCount,
+            -1)
+    df$span_cnt <-
+        ifelse(!is.na(sample_file$SpanningFragCount),
+            sample_file$SpanningFragCount,
+            -1)
+    df$t_area1 <- sample_file$SpliceType
+    df$t_area2 <- sample_file$SpliceType
+
+    return(df)
+
+}
+
+sample_file <- fread(opt[2], data.table = F)
+
+tool_cff <-
+    setNames(data.frame(matrix(
+    ncol = length(cff_format), nrow = 0
+    )), cff_format)
+if (opt[1] == "arriba") {
+    if (nrow(sample_file) > 0) {
+        tool_cff <- make_arriba(sample_file)
+    }
+} else if (opt[1]  == "fusioncatcher") {
+    if (nrow(sample_file) > 0) {
+        tool_cff <- make_fusioncatcher(sample_file)
+    }
+} else if (opt[1]  == "starfusion") {
+    if (nrow(sample_file) > 0) {
+        tool_cff <- make_starfusion(sample_file)
+    }
+}
+
+if (nrow(tool_cff) > 0) {
+    tool_cff$sample_name <- opt[3]
+    tool_cff$library <- "RNA"
+    tool_cff$sample_type <- "Tumor"
+    tool_cff$disease <- NA
+}
+tool_cff$strand1[tool_cff$strand1 == "."] <- NA
+tool_cff$strand2[tool_cff$strand2 == "."] <- NA
+tool_cff <- tool_cff[, cff_format]
+write.table(
+    tool_cff,
+    opt[4],
+    sep = "\t",
+    row.names = F,
+    quote =  F,
+    col.names = F
+)
diff --git a/bin/make_gene_info_for_forte.R b/bin/make_gene_info_for_forte.R
new file mode 100644
index 0000000..1ef403c
--- /dev/null
+++ b/bin/make_gene_info_for_forte.R
@@ -0,0 +1,83 @@
+
+# __author__      = "Alexandria Dymun"
+# __email__       = "pintoa1@mskcc.org"
+# __contributor__ = "Anne Marie Noronha (noronhaa@mskcc.org)"
+# __version__     = "0.0.1"
+# __status__      = "Dev"
+
+
+
+library(dplyr)
+library(stringr)
+library(argparse)
+
+opt = commandArgs(TRUE)
+
+parser=ArgumentParser()
+parser$add_argument("-p",'--primary_gtf',type="character",default = NULL,help = "Primary GTF, should match your bed file and arriba. Assumes ARRIBA is on primary gtf")
+parser$add_argument("-c",'--fc_custom_bed_gene_names',type="character",default = NULL,help = "Fusioncatcher custom genes bed file")
+parser$add_argument("-s",'--star_fusion_ref',type="character",default = NULL,help = "StarFusion GTF")
+parser$add_argument("-f",'--fusioncatcher_ref',type="character",default = NULL,help = "Fusioncatcher GTF")
+parser$add_argument("-o",'--outputDir',type="character",default = NULL,help = "outputDirectory to write gene_info and excess gene list")
+
+opt=parser$parse_args()
+
+
+### primary gtf is v75, also used in arriba
+primary_gtf <-  as.data.frame(rtracklayer::import(opt$primary_gtf))
+
+### Fusion catcher has custom gene names/gene_ids....
+## https://github.com/ndaniel/fusioncatcher/blob/ebc46fd1a8046fc909a56e09944a2ec2d69cc808/bin/add_custom_gene.py#L704-L715
+fc_custom_bed_gene_names <- read.table(opt$fc_custom_bed_gene_names)
+fc_custom_bed_gene_names$gene_name <- str_split_fixed(fc_custom_bed_gene_names$V4,"-",n=2)[,1]
+fc_custom_bed_gene_names$gene_id <- str_split_fixed(fc_custom_bed_gene_names$V4,"-",n=3)[,2]
+star_fusion_ref <- as.data.frame(rtracklayer::import(opt$star_fusion_ref))
+fusioncatcher_ref <- as.data.frame(rtracklayer::import(opt$fusioncatcher_ref))
+
+
+all_my_gene_ids_and_names <- list(primary_gtf,fc_custom_bed_gene_names,star_fusion_ref,fusioncatcher_ref)
+### whichever gtf you label as primary should also be the same version of the gene_bed file generated for metafusion
+names(all_my_gene_ids_and_names) <- c("primary","one","two","three")
+unique_id_to_names <- lapply( all_my_gene_ids_and_names,function(gtf) {
+    ### If gene id has versions, strip them off
+    if(all(grepl("\\.",gtf$gene_id))){
+        gtf$gene_id_with_version <- gtf$gene_id
+        gtf$gene_id <- str_split_fixed(gtf$gene_id_with_version,"\\.",n=2)[,1]
+        return(unique(gtf[,c("gene_name","gene_id","gene_id_with_version")]))
+    } else{
+        return(unique(gtf[,c("gene_name","gene_id")]))
+    }
+})
+
+
+gene_info <- unique_id_to_names$primary
+### tack on missing gene_ids from other references to gene info
+versioned_gtf <-unlist(sapply(names(unique_id_to_names)[names(unique_id_to_names) != "primary"],function(name){
+    if(any(colnames(unique_id_to_names[[name]]) == "gene_id_with_version")){
+    return(name)
+    }
+}))
+
+
+add_these_exess_gene_ids <- do.call(rbind,lapply(names(unique_id_to_names)[names(unique_id_to_names) != "primary"],function(name){
+    add_symbols_and_ids <- unique_id_to_names[[name]]
+    add_symbols_and_ids <- add_symbols_and_ids[!add_symbols_and_ids$gene_id %in% gene_info$gene_id,]
+    if(name %in% versioned_gtf){
+    add_symbols_and_ids <-add_symbols_and_ids[,c("gene_name","gene_id_with_version")]
+    colnames(add_symbols_and_ids) <- c("gene_name","gene_id")
+    }
+    return(add_symbols_and_ids)
+
+}))
+# Excess genes being added (genes will be flagged as gene not in v75)
+gene_info <- rbind(gene_info,add_these_exess_gene_ids)
+
+gene_info <- merge(gene_info,do.call(rbind,unique_id_to_names[versioned_gtf])[,c("gene_id","gene_id_with_version")],by = "gene_id",all.x = T, all.y = F)
+
+gene_info$Synonyms <- ifelse(is.na(gene_info$gene_id_with_version),gene_info$gene_id,paste0(gene_info$gene_id,"|",gene_info$gene_id_with_version))
+gene_info$Symbol <- gene_info$gene_name
+
+gene_info <- gene_info[,c("Symbol","Synonyms")]
+
+write.table(gene_info,paste0(opt$outputDir,"/gene.info"),sep ="\t",quote = F,row.names = F)
+write.table(add_these_exess_gene_ids,paste0(opt$outputDir,"/excess_gene_ids.txt",sep ="\t",quote = F,row.names = F)
diff --git a/conf/igenomes.config b/conf/igenomes.config
index d734a1b..a85e19d 100644
--- a/conf/igenomes.config
+++ b/conf/igenomes.config
@@ -33,6 +33,10 @@ params {
                     baits = "${params.targets_base}/${params.genome}/baits/agilent_v4/Agilent_v4_51MB_Human_b37_baits.bed"
                 }
             }
+            metafusion_blocklist = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37/blocklist_breakpoints.bedpe.gz"
+            metafusion_gene_bed  = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37/v75_gene.bed.gz"
+            metafusion_gene_info = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37/gene_info_20230714.txt"
+            ensembl_version = 75
         }
         'GRCh38' {
             fasta          = "${params.igenomes_base}/Homo_sapiens/NCBI/GRCh38/Sequence/WholeGenomeFasta/genome.fa"
@@ -52,7 +56,12 @@ params {
             arriba_blacklist       = "/usr/local/var/lib/arriba/blacklist_hg19_hs37d5_GRCh37_v2.3.0.tsv.gz"
             arriba_known_fusions   = "/usr/local/var/lib/arriba/known_fusions_hg19_hs37d5_GRCh37_v2.3.0.tsv.gz"
             arriba_protein_domains = "/usr/local/var/lib/arriba/protein_domains_hg19_hs37d5_GRCh37_v2.3.0.gff3"
-            cdna           = "https://ftp.ensembl.org/pub/release-75/fasta/homo_sapiens/cdna/Homo_sapiens.GRCh37.75.cdna.all.fa.gz"
+            cdna           = "http://ftp.ensemblgenomes.org/pub/viruses/fasta/sars_cov_2/cdna/Sars_cov_2.ASM985889v3.cdna.all.fa.gz"
+            metafusion_blocklist = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37_test/blocklist_breakpoints.bedpe"
+            metafusion_gene_bed  = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37_test/v75_gene.bed"
+            metafusion_gene_info = "https://raw.githubusercontent.com/anoronh4/forte-references/main/GRCh37_test/gene_info_20230714.txt"
+            ensembl_version = 75
+
         }
 /*
         'hg38' {
diff --git a/conf/modules.config b/conf/modules.config
index 0927b20..2be37bf 100644
--- a/conf/modules.config
+++ b/conf/modules.config
@@ -171,22 +171,39 @@ process {
         ext.args = "--cosmic_usr ${params.cosmic_usr} --cosmic_passwd \$COSMIC_PASSWD"
     }
 
-    withName: FUSIONREPORT {
-        ext.when = ! params.skip_fusion_report
-        ext.args = {
-            [
-                "-t ${params.fusion_report_cutoff}",
-                '--allow-multiple-gene-symbols',
-                '--export csv'
-            ].join(' ').trim()
-        }
+    withName: MERGE_CFF {
+        publishDir = [
+            enabled: false
+        ]
+    }
+
+    withName: METAFUSION {
+        ext.args = "--num_tools=${params.fusion_tool_cutoff}"
+        publishDir = [
+            path: { "$params.outdir/analysis/${meta.id}/metafusion/intermediates" },
+            mode: params.publish_dir_mode,
+            enabled: true,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+
+        ]
+    }
+
+    withName: ADD_FLAG {
+        publishDir = [
+            path: { "$params.outdir/analysis/${meta.id}/metafusion/intermediates" },
+            mode: params.publish_dir_mode,
+            enabled: true,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : filename }
+        ]
     }
 
-    withName: CSV_TO_TSV {
+    withName: AGFUSION_BATCH {
+        ext.args = "-a cff_transcript --middlestar"
         publishDir = [
-            path: { "${params.outdir}/analysis/${meta.id}/fusionreport" },
+            path: { "${params.outdir}/analysis/${meta.id}/agfusion" },
             mode: params.publish_dir_mode,
-            pattern: '*.tsv'
+            pattern: 'fusion_transcripts.csv',
+            enabled: false
         ]
     }
 
@@ -194,6 +211,22 @@ process {
         ext.when = params.run_oncokb_fusionannotator
         secret = ["ONCOKB_TOKEN"]
         ext.args = "-b \$ONCOKB_TOKEN"
+        publishDir = [
+            path: { "${params.outdir}/analysis/${meta.id}/oncokb" },
+            mode: params.publish_dir_mode,
+            pattern: '*.oncokb.tsv',
+            enabled: false
+        ]
+    }
+
+    withName: CFF_FINALIZE {
+        ext.prefix = { "${meta.id}" }
+        publishDir = [
+            path: { "${params.outdir}/analysis/${meta.id}/metafusion" },
+            mode: params.publish_dir_mode,
+            pattern: '*.{final,unfiltered}.cff',
+            enabled: true
+        ]
     }
 
     withName: STAR_FOR_ARRIBA {
diff --git a/main.nf b/main.nf
index 8a4750e..5dfe5f6 100644
--- a/main.nf
+++ b/main.nf
@@ -24,6 +24,10 @@ params.cdna           = WorkflowMain.getGenomeAttribute(params, 'cdna')
 params.arriba_blacklist       = WorkflowMain.getGenomeAttribute(params, 'arriba_blacklist')
 params.arriba_known_fusions   = WorkflowMain.getGenomeAttribute(params, 'arriba_known_fusions')
 params.arriba_protein_domains = WorkflowMain.getGenomeAttribute(params, 'arriba_protein_domains')
+params.metafusion_blocklist = WorkflowMain.getGenomeAttribute(params, 'metafusion_blocklist')
+params.metafusion_gene_bed  = WorkflowMain.getGenomeAttribute(params, 'metafusion_gene_bed')
+params.metafusion_gene_info = WorkflowMain.getGenomeAttribute(params, 'metafusion_gene_info')
+params.ensembl_version      = WorkflowMain.getGenomeAttribute(params, 'ensembl_version')
 
 WorkflowMain.initialise(workflow, params, log)
 
diff --git a/modules.json b/modules.json
index 2c0b137..9228792 100644
--- a/modules.json
+++ b/modules.json
@@ -10,12 +10,12 @@
                         "git_sha": "c8e35eb2055c099720a75538d1b8adb3fb5a464c",
                         "installed_by": ["modules"]
                     },
-                    "cat/fastq": {
+                    "cat/cat": {
                         "branch": "master",
-                        "git_sha": "c8e35eb2055c099720a75538d1b8adb3fb5a464c",
+                        "git_sha": "911696ea0b62df80e900ef244d7867d177971f73",
                         "installed_by": ["modules"]
                     },
-                    "csvtk/concat": {
+                    "cat/fastq": {
                         "branch": "master",
                         "git_sha": "c8e35eb2055c099720a75538d1b8adb3fb5a464c",
                         "installed_by": ["modules"]
diff --git a/modules/local/add_flags/main.nf b/modules/local/add_flags/main.nf
new file mode 100644
index 0000000..1e05dbc
--- /dev/null
+++ b/modules/local/add_flags/main.nf
@@ -0,0 +1,37 @@
+process ADD_FLAG {
+    tag "$meta.id"
+    label "process_single"
+
+/// must be using singularity 3.7+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' :
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' }"
+
+    input:
+    tuple val(meta), path(cluster), path(cis), path(cff), path(problem_chrom), path(filters)
+
+    output:
+    tuple val(meta), path("*_metafusion_cluster.unfiltered.cff"), emit: unfiltered_cff
+    path "versions.yml"                                         , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def sample = "${meta.sample}"
+    """
+    add_flags_and_cluster_information.R \\
+        $cff \\
+        $cluster \\
+        $cis \\
+        $problem_chrom \\
+        $filters \\
+        $sample
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        R: \$(R --version | head -n1)
+        add_flags_and_cluster_information.R: 0.0.1
+    END_VERSIONS
+    """
+}
diff --git a/modules/local/agfusion/batch/main.nf b/modules/local/agfusion/batch/main.nf
new file mode 100644
index 0000000..c4dd105
--- /dev/null
+++ b/modules/local/agfusion/batch/main.nf
@@ -0,0 +1,47 @@
+process AGFUSION_BATCH {
+    tag "$meta.id"
+    label 'process_low'
+
+    // Note: 2.7X indices incompatible with AWS iGenomes.
+    conda 'bioconda::agfusion=1.252'
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'cmopipeline/agfusion:0.0.5' :
+        'cmopipeline/agfusion:0.0.5' }"
+
+    input:
+    tuple val(meta), path(fusions)
+    path(agfusion_db)
+    path(pyensembl_cache)
+
+    output:
+    tuple val(meta), path("${prefix}")                       , emit: fusions_annotated
+    tuple val(meta), path("${prefix}.fusion_transcripts.csv"), emit: fusion_transcripts_csv
+    tuple val(meta), path("${prefix}.fusion_transcripts.tsv"), emit: fusion_transcripts_tsv
+    path "versions.yml"                                      , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args   = task.ext.args ?: ''
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    export PYENSEMBL_CACHE_DIR=\$PWD/${pyensembl_cache}
+
+    awk -F"\\t" 'NR == 1 && \$1 ~ /gene5_chr/ {next;}{print}' ${fusions} > ${fusions}.no_header
+
+    agfusion batch \\
+        -f ${fusions}.no_header \\
+        -db ${agfusion_db} \\
+        -o ${prefix} \\
+        ${args}
+
+    awk -F"," 'NR != 1 && FNR == 1 {next;}{print}' ${prefix}/*/*.fusion_transcripts.csv > ${prefix}.fusion_transcripts.csv
+    cat ${prefix}.fusion_transcripts.csv | tr "," "\\t" > ${prefix}.fusion_transcripts.tsv
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        agfusion: \$(agfusion -v) (fork)
+    END_VERSIONS
+    """
+}
diff --git a/modules/local/agfusion/batch/meta.yml b/modules/local/agfusion/batch/meta.yml
new file mode 100644
index 0000000..7961454
--- /dev/null
+++ b/modules/local/agfusion/batch/meta.yml
@@ -0,0 +1,49 @@
+name: agfusion_batch
+description: AGFusion batch annotation
+keywords:
+  - agfusion
+  - fusion
+  - batch
+  - frame
+tools:
+  - batch:
+      description: Annotate a file containing fusions
+      homepage: https://github.com/anoronh4/AGFusion
+      documentation: https://github.com/anoronh4/AGFusion/blob/master/README.md
+
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - fusions:
+      type: path
+      description: Fusions in one of many formats (may include arriba outputs or cff)
+      pattern: "*"
+  - agfusion_db:
+      type: path
+      description: File containing AGFusion reference information.
+      pattern: "*.db"
+  - pyensembl_cache:
+      type: path
+      description: Folder containing pyensembl cache.
+      pattern: "*"
+
+output:
+  - versions:
+      type: path
+      description: File containing software versions
+      pattern: "versions.yml"
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - agfusion_result:
+      type: path
+      description: Folder containing annotations
+      pattern: "${prefix}/"
+
+authors:
+  - "@anoronh4"
diff --git a/modules/local/agfusion/container/Dockerfile b/modules/local/agfusion/container/Dockerfile
new file mode 100755
index 0000000..3c20dd9
--- /dev/null
+++ b/modules/local/agfusion/container/Dockerfile
@@ -0,0 +1,26 @@
+FROM ubuntu:bionic-20230530
+
+LABEL maintainer="Anne Marie Noronha (noronhaa@mskcc.org)" \
+    version.image="0.0.5"
+
+# INSTALL DEPENDENCIES
+
+ENV DEBIAN_FRONTEND=noninteractive
+
+RUN apt-get update -y
+RUN apt-get install -y build-essential python3 python3-pip python3-matplotlib python3-pandas python3-future python3-biopython curl less vim libnss-sss git zip
+RUN pip3 install --upgrade pip
+RUN pip3 install pyensembl
+
+# Additional libraries needed for AGFusion build command
+RUN apt-get install -y default-libmysqlclient-dev
+RUN pip3 install mysqlclient
+
+# INSTALL AGFUSION & DATABASE FILES
+WORKDIR /usr/local/bin
+RUN git clone https://github.com/anoronh4/AGFusion.git --branch feature/metafusion_parser
+WORKDIR /usr/local/bin/AGFusion
+RUN pip3 install .
+
+# downgrade pyensembl for compatibility
+RUN pip3 install gtfparse==1.2.1 --upgrade
diff --git a/modules/local/agfusion/download/main.nf b/modules/local/agfusion/download/main.nf
new file mode 100644
index 0000000..5744641
--- /dev/null
+++ b/modules/local/agfusion/download/main.nf
@@ -0,0 +1,58 @@
+process AGFUSION_DOWNLOAD {
+    label 'process_low'
+
+    // Note: 2.7X indices incompatible with AWS iGenomes.
+    conda 'bioconda::agfusion=1.252'
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'cmopipeline/agfusion:0.0.5' :
+        'cmopipeline/agfusion:0.0.5' }"
+
+    input:
+    val(ensembl_release)
+    val(genome)
+
+    output:
+    path "agfusion.*.db"  , emit: agfusion_db
+    path "pyensembl_cache", emit: pyensembl_cache
+    path "versions.yml"   , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def agfusion_genome = ['GRCh37','smallGRCh37','hg19'].contains(genome) ? 'hg19' :
+        ['GRCh38','hg38'].contains(genome) ? 'hg38' :
+        ['GRCm38','mm10'].contains(genome) ? 'mm10' : ''
+    def pyensembl_species = ['GRCm38','mm10'].contains(genome) ? 'mus_musculus' : 'homo_sapiens'
+    if (ensembl_release < 93) {
+        """
+        export PYENSEMBL_CACHE_DIR=\$PWD/pyensembl_cache
+
+        pyensembl install --species ${pyensembl_species} --release ${ensembl_release}
+
+        agfusion download -g ${agfusion_genome}
+
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            agfusion: \$(agfusion -v) (fork)
+        END_VERSIONS
+        """
+    } else {
+        """
+        export PYENSEMBL_CACHE_DIR=\$PWD/pyensembl_cache
+
+        pyensembl install --species ${pyensembl_species} --release ${ensembl_release}
+
+        curl http://ftp.ebi.ac.uk/pub/databases/Pfam/current_release/database_files/pfamA.txt.gz > pfamA.txt.gz
+        gunzip pfamA.txt.gz
+        agfusion build --dir . --species ${agfusion_genome} --release ${ensembl_release} --pfam pfamA.txt
+        rm pfamA.txt
+
+        cat <<-END_VERSIONS > versions.yml
+        "${task.process}":
+            agfusion: \$(agfusion -v) (fork)
+        END_VERSIONS
+        """
+    }
+}
diff --git a/modules/local/agfusion/download/meta.yml b/modules/local/agfusion/download/meta.yml
new file mode 100644
index 0000000..21a15a8
--- /dev/null
+++ b/modules/local/agfusion/download/meta.yml
@@ -0,0 +1,35 @@
+name: fusionreport_download
+description: Build DB for fusionreport
+keywords:
+  - sort
+tools:
+  - fusioncatcher:
+      description: Build DB for fusionreport
+      homepage: https://github.com/ndaniel/fusioncatcher/
+      documentation: https://github.com/ndaniel/fusioncatcher/blob/master/doc/manual.md
+      tool_dev_url: https://github.com/ndaniel/fusioncatcher/
+      doi: "10.1101/011650"
+      licence: ["GPL v3"]
+
+input:
+  - username:
+      type: value
+      description: Organism for which the data is downloaded from Ensembl database and built
+      pattern: "*"
+  - passwd:
+      type: value
+      description: Organism for which the data is downloaded from Ensembl database and built
+      pattern: "*"
+
+output:
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - reference:
+      type: directory
+      description: directory containing the genome resource files required for fusioncatcher
+      pattern: "fusioncatcher-genome"
+
+authors:
+  - "@praveenraj2018"
diff --git a/modules/local/cff_annotate/main.nf b/modules/local/cff_annotate/main.nf
new file mode 100644
index 0000000..dc4a16b
--- /dev/null
+++ b/modules/local/cff_annotate/main.nf
@@ -0,0 +1,35 @@
+process CFF_ANNOTATE {
+    tag "$meta.id"
+    label "process_single"
+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' :
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' }"
+
+    input:
+    tuple val(meta), path(cff), path(oncokb), path(agfusion)
+
+    output:
+    tuple val(meta), path("${prefix}.unfiltered.cff"), emit: unfiltered_cff
+    tuple val(meta), path("${prefix}.final.cff")     , emit: filtered_cff
+    path "versions.yml"                              , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    prefix = task.ext.prefix ?: "${cff}"
+    """
+    add_annotations_cff.R \\
+        --cff ${cff} \\
+        --oncokb ${oncokb} \\
+        --agfusion ${agfusion} \\
+        --out-prefix ${prefix}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        R: \$(R --version | head -n1)
+        add_annotations_cff.R: 0.0.1
+    END_VERSIONS
+    """
+}
diff --git a/modules/local/convert_to_cff/main.nf b/modules/local/convert_to_cff/main.nf
new file mode 100644
index 0000000..ec5ecf0
--- /dev/null
+++ b/modules/local/convert_to_cff/main.nf
@@ -0,0 +1,31 @@
+process TO_CFF {
+    tag "$meta.id"
+    label "process_single"
+
+//  must be using singularity 3.7+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' :
+        'ghcr.io/rocker-org/devcontainer/tidyverse:4' }"
+
+    input:
+    tuple val(meta), val(caller), path(fusions)
+
+    output:
+    tuple val(meta), path("*.cff"), emit: cff
+    path "versions.yml"           , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def sample = "${meta.sample}"
+    """
+    make_cff_from_forte.R $caller $fusions $sample ${sample}_${caller}.cff
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        R: \$(R --version | head -n1)
+        make_cff_from_forte.R: 0.0.1
+    END_VERSIONS
+    """
+}
diff --git a/modules/local/metafusion/container/Dockerfile b/modules/local/metafusion/container/Dockerfile
new file mode 100755
index 0000000..018cef8
--- /dev/null
+++ b/modules/local/metafusion/container/Dockerfile
@@ -0,0 +1,23 @@
+FROM mapostolides/metafusion
+
+LABEL author="Alexandria Dymun (pintoa1@mskcc.org)" \
+    maintainer="Anne Marie Noronha (noronhaa@mskcc.org)" \
+    version.image="0.0.6"
+
+ENV METAFUSION_TAG="v1.0.1"
+ENV PATH="${PATH}:/MetaFusion/scripts"
+
+RUN R -e "chooseCRANmirror(ind=52); install.packages(c('plyr','data.table'))"
+
+RUN git clone -b $METAFUSION_TAG https://github.com/pintoa1-mskcc/MetaFusion.git --single-branch && \
+    cd /MetaFusion/ && \
+    git pull && \
+    chmod -R +x /MetaFusion/scripts && \
+    rm -rf /MetaFusion/test_data
+
+# run cleanup
+RUN conda clean -afy \
+    && find /opt/conda/ -follow -type f -name '*.a' -delete \
+    && find /opt/conda/ -follow -type f -name '*.pyc' -delete \
+    && find /opt/conda/ -follow -type f -name '*.js.map' -delete
+
diff --git a/modules/local/metafusion/main.nf b/modules/local/metafusion/main.nf
new file mode 100644
index 0000000..e3e8120
--- /dev/null
+++ b/modules/local/metafusion/main.nf
@@ -0,0 +1,47 @@
+process METAFUSION {
+    tag "$meta.id"
+    label "process_low"
+
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'cmopipeline/metafusion:0.0.6' :
+        'cmopipeline/metafusion:0.0.6' }"
+
+    input:
+    tuple val(meta), path(cff)
+    path genebed
+    path info
+    path fasta
+    path blocklist
+
+    output:
+    tuple val(meta), path("*final*cluster")             , emit: cluster
+    tuple val(meta), path("*.exons")                    , emit: cff
+    tuple val(meta), path("cis-sage.cluster")           , emit: cis
+    tuple val(meta), path("problematic_chromosomes.cff"), emit: problem_chrom
+    tuple val(meta), path("filters.txt")                , emit: filters
+    // tuple val(meta), path("*")                       , emit: all
+    path "versions.yml"                                 , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    args = task.ext.args ?: ""
+    def sample = "${meta.sample}"
+    """
+    Metafusion_forte.sh \\
+        --cff $cff \\
+        --outdir . \\
+        --gene_bed $genebed \\
+        --gene_info $info \\
+        --genome_fasta $fasta \\
+        --recurrent_bedpe $blocklist \\
+        ${args}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        Metafusion docker: \$METAFUSION_TAG
+        Metafusion_forte.sh: 0.0.1
+    END_VERSIONS
+    """
+}
diff --git a/modules/local/oncokb/fusionannotator/main.nf b/modules/local/oncokb/fusionannotator/main.nf
index 16720e0..1a263c5 100644
--- a/modules/local/oncokb/fusionannotator/main.nf
+++ b/modules/local/oncokb/fusionannotator/main.nf
@@ -3,16 +3,16 @@ process ONCOKB_FUSIONANNOTATOR {
     label 'process_low'
 
     // Note: 2.7X indices incompatible with AWS iGenomes.
-    conda "shahcompbio::oncokb-annotator=2.3.3"
+    //conda "shahcompbio::oncokb-annotator=2.3.3"
     container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
         'cmopipeline/oncokb-annotator:0.0.1' :
         'cmopipeline/oncokb-annotator:0.0.1' }"
 
     input:
-    tuple val(meta), path(fusions)
+    tuple val(meta), path(cff)
 
     output:
-    tuple val(meta), path("*.oncokb.tsv"), emit: oncokb_fusions
+    tuple val(meta), path("*.oncokb.tsv")        , emit: oncokb_fusions
     path "versions.yml"                          , emit: versions
 
     when:
@@ -23,9 +23,10 @@ process ONCOKB_FUSIONANNOTATOR {
     def args2 = task.ext.args2 ?: ''
     def prefix = task.ext.prefix ?: "${meta.id}"
     """
-    awk 'BEGIN {FS=OFS="|"}{gsub("--","-",\$1)}1' ${fusions} > ${fusions}.reformat
+    awk 'BEGIN { FS = "\\t"; OFS = "\\t"} {print \$31,\$19"-"\$21}' ${cff} | tail -n+2 > ${cff}.reformat
+    echo -e "Tumor_Sample_Barcode\tFusion" | cat - ${cff}.reformat  > ${cff}.1reformat
     FusionAnnotator.py \\
-        -i ${fusions}.reformat \\
+        -i ${cff}.1reformat \\
         -o ${prefix}.oncokb.tsv \\
         ${args}
 
diff --git a/modules/nf-core/cat/cat/main.nf b/modules/nf-core/cat/cat/main.nf
new file mode 100644
index 0000000..9f06221
--- /dev/null
+++ b/modules/nf-core/cat/cat/main.nf
@@ -0,0 +1,62 @@
+process CAT_CAT {
+    tag "$meta.id"
+    label 'process_low'
+
+    conda "conda-forge::pigz=2.3.4"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/pigz:2.3.4' :
+        'biocontainers/pigz:2.3.4' }"
+
+    input:
+    tuple val(meta), path(files_in)
+
+    output:
+    tuple val(meta), path("${prefix}"), emit: file_out
+    path "versions.yml"               , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+    def args2 = task.ext.args2 ?: ''
+    def file_list = files_in.collect { it.toString() }
+
+    // | input     | output     | command1 | command2 |
+    // |-----------|------------|----------|----------|
+    // | gzipped   | gzipped    | cat      |          |
+    // | ungzipped | ungzipped  | cat      |          |
+    // | gzipped   | ungzipped  | zcat     |          |
+    // | ungzipped | gzipped    | cat      | pigz     |
+
+    // Use input file ending as default
+    prefix   = task.ext.prefix ?: "${meta.id}${file_list[0].substring(file_list[0].lastIndexOf('.'))}"
+    out_zip  = prefix.endsWith('.gz')
+    in_zip   = file_list[0].endsWith('.gz')
+    command1 = (in_zip && !out_zip) ? 'zcat' : 'cat'
+    command2 = (!in_zip && out_zip) ? "| pigz -c -p $task.cpus $args2" : ''
+    """
+    $command1 \\
+        $args \\
+        ${file_list.join(' ')} \\
+        $command2 \\
+        > ${prefix}
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        pigz: \$( pigz --version 2>&1 | sed 's/pigz //g' )
+    END_VERSIONS
+    """
+
+    stub:
+    def file_list = files_in.collect { it.toString() }
+    prefix   = task.ext.prefix ?: "${meta.id}${file_list[0].substring(file_list[0].lastIndexOf('.'))}"
+    """
+    touch $prefix
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        pigz: \$( pigz --version 2>&1 | sed 's/pigz //g' )
+    END_VERSIONS
+    """
+}
diff --git a/modules/nf-core/cat/cat/meta.yml b/modules/nf-core/cat/cat/meta.yml
new file mode 100644
index 0000000..8acc0bf
--- /dev/null
+++ b/modules/nf-core/cat/cat/meta.yml
@@ -0,0 +1,37 @@
+name: cat_cat
+description: A module for concatenation of gzipped or uncompressed files
+keywords:
+  - concatenate
+  - gzip
+  - cat
+tools:
+  - cat:
+      description: Just concatenation
+
+      documentation: https://man7.org/linux/man-pages/man1/cat.1.html
+
+      licence: ["GPL-3.0-or-later"]
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [ id:'test', single_end:false ]
+  - files_in:
+      type: file
+      description: List of compressed / uncompressed files
+      pattern: "*"
+
+output:
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+  - file_out:
+      type: file
+      description: Concatenated file. Will be gzipped if file_out ends with ".gz"
+      pattern: "${file_out}"
+
+authors:
+  - "@erikrikarddaniel"
+  - "@FriederikeHanssen"
diff --git a/modules/nf-core/csvtk/concat/main.nf b/modules/nf-core/csvtk/concat/main.nf
deleted file mode 100644
index c9fb9bf..0000000
--- a/modules/nf-core/csvtk/concat/main.nf
+++ /dev/null
@@ -1,43 +0,0 @@
-process CSVTK_CONCAT {
-    tag "$meta.id"
-    label 'process_low'
-
-    conda "bioconda::csvtk=0.23.0"
-    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
-        'https://depot.galaxyproject.org/singularity/csvtk:0.23.0--h9ee0642_0' :
-        'quay.io/biocontainers/csvtk:0.23.0--h9ee0642_0' }"
-
-    input:
-    tuple val(meta), path(csv)
-    val in_format
-    val out_format
-
-    output:
-    tuple val(meta), path("${prefix}.${out_extension}"), emit: csv
-    path "versions.yml"                                , emit: versions
-
-    when:
-    task.ext.when == null || task.ext.when
-
-    script:
-    def args = task.ext.args   ?: ''
-    prefix   = task.ext.prefix ?: "${meta.id}"
-    def delimiter = in_format == "tsv" ? "\t" : (in_format == "csv" ? "," : in_format)
-    def out_delimiter = out_format == "tsv" ? "\t" : (out_format == "csv" ? "," : out_format)
-    out_extension = out_format == "tsv" ? 'tsv' : 'csv'
-    """
-    csvtk \\
-        concat \\
-        $args \\
-        --num-cpus $task.cpus \\
-        --delimiter "${delimiter}" \\
-        --out-delimiter "${out_delimiter}" \\
-        --out-file ${prefix}.${out_extension} \\
-        $csv
-
-    cat <<-END_VERSIONS > versions.yml
-    "${task.process}":
-        csvtk: \$(echo \$( csvtk version | sed -e "s/csvtk v//g" ))
-    END_VERSIONS
-    """
-}
diff --git a/modules/nf-core/csvtk/concat/meta.yml b/modules/nf-core/csvtk/concat/meta.yml
deleted file mode 100644
index 2d2f856..0000000
--- a/modules/nf-core/csvtk/concat/meta.yml
+++ /dev/null
@@ -1,51 +0,0 @@
-name: csvtk_concat
-description: Concatenate two or more CSV (or TSV) tables into a single table
-keywords:
-  - concatenate
-  - tsv
-  - csv
-tools:
-  - csvtk:
-      description: A cross-platform, efficient, practical CSV/TSV toolkit
-      homepage: http://bioinf.shenwei.me/csvtk
-      documentation: http://bioinf.shenwei.me/csvtk
-      tool_dev_url: https://github.com/shenwei356/csvtk
-      doi: ""
-      licence: ["MIT"]
-
-input:
-  - meta:
-      type: map
-      description: |
-        Groovy Map containing sample information
-        e.g. [ id:'test', single_end:false ]
-  - csv:
-      type: file
-      description: CSV/TSV formatted files
-      pattern: "*.{csv,tsv}"
-  - in_format:
-      type: string
-      description: Input format (csv, tab, or a delimiting character)
-      pattern: "*"
-  - out_format:
-      type: string
-      description: Output format (csv, tab, or a delimiting character)
-      pattern: "*"
-
-output:
-  - meta:
-      type: map
-      description: |
-        Groovy Map containing sample information
-        e.g. [ id:'test', single_end:false ]
-  - versions:
-      type: file
-      description: File containing software versions
-      pattern: "version.yml"
-  - csv:
-      type: file
-      description: Concatenated CSV/TSV file
-      pattern: "*.{csv,tsv}"
-
-authors:
-  - "@rpetit3"
diff --git a/nextflow.config b/nextflow.config
index 234d21b..5c026f9 100644
--- a/nextflow.config
+++ b/nextflow.config
@@ -33,6 +33,7 @@ params {
     cosmic_usr                 = null
     fusion_report_cutoff       = 1
     skip_fusion_report         = false
+    fusion_tool_cutoff         = 1
 
     // QC
     // rseqc_modules can include ['bam_stat','inner_distance','infer_experiment','junction_annotation','junction_saturation','read_distribution','read_duplication','tin']
diff --git a/subworkflows/local/fusion.nf b/subworkflows/local/fusion.nf
index 4f045ee..af9c7cc 100644
--- a/subworkflows/local/fusion.nf
+++ b/subworkflows/local/fusion.nf
@@ -3,9 +3,15 @@ include { ARRIBA                            } from '../../modules/nf-core/arriba
 include { STAR_ALIGN as STAR_FOR_STARFUSION } from '../../modules/nf-core/star/align/main'
 include { STARFUSION                        } from '../../modules/local/starfusion/detect/main'
 include { FUSIONCATCHER_DETECT              } from '../../modules/local/fusioncatcher/detect/main'
-include { FUSIONREPORT                      } from '../../modules/local/fusionreport/run/main'
 include { ONCOKB_FUSIONANNOTATOR            } from '../../modules/local/oncokb/fusionannotator/main'
-include { CSVTK_CONCAT as CSV_TO_TSV        } from '../../modules/nf-core/csvtk/concat/main'
+include { AGFUSION_BATCH                    } from '../../modules/local/agfusion/batch/main'
+include { TO_CFF as ARRIBA_TO_CFF           } from '../../modules/local/convert_to_cff/main'
+include { TO_CFF as FUSIONCATCHER_TO_CFF    } from '../../modules/local/convert_to_cff/main'
+include { TO_CFF as STARFUSION_TO_CFF       } from '../../modules/local/convert_to_cff/main'
+include { CAT_CAT as MERGE_CFF              } from '../../modules/nf-core/cat/cat/main'
+include { METAFUSION                        } from '../../modules/local/metafusion/main'
+include { ADD_FLAG                          } from '../../modules/local/add_flags/main'
+include { CFF_ANNOTATE as CFF_FINALIZE      } from '../../modules/local/cff_annotate/main'
 
 workflow FUSION {
 
@@ -16,10 +22,17 @@ workflow FUSION {
     starfusion_ref
     fusioncatcher_ref
     fusion_report_db
+    agfusion_db
+    pyensembl_cache
+    gene_bed
+    blocklist
 
     main:
     ch_versions = Channel.empty()
     fasta = params.fasta
+    //gene_bed = params.metafusion_gene_bed
+    gene_info = params.metafusion_gene_info
+    //blocklist = params.metafusion_blocklist
 
     STAR_FOR_ARRIBA(
         reads,
@@ -70,42 +83,65 @@ workflow FUSION {
 
     fc_fusions = ["GRCh37","hg19","smallGRCh37"].contains(params.genome) ? FUSIONCATCHER_DETECT.out.fusions_alt : FUSIONCATCHER_DETECT.out.fusions
 
+
+    ARRIBA_TO_CFF(ARRIBA.out.fusions
+            .map{ meta, file ->[ meta, "arriba", file ] })
+    FUSIONCATCHER_TO_CFF(fc_fusions
+                    .map{ meta, file -> [ meta, "fusioncatcher", file ] } )
+    STARFUSION_TO_CFF(STARFUSION.out.abridged
+                    .map{ meta, file -> [ meta, "starfusion", file ] })
     // get expected number of callers for groupTuple
     numcallers = 1 + ( params.starfusion_url ? 1 : 0 ) + ( ["GRCh37","GRCh38"].contains(params.genome) ? 1 : 0 )
 
-    FUSIONREPORT(
-        ARRIBA.out.fusions
-            .map{ meta, file ->[ meta, "arriba", file ] }
+    MERGE_CFF(
+        ARRIBA_TO_CFF.out.cff
+            .map{ meta, file -> [meta, file]}
             .mix(
-                fc_fusions
-                    .map{ meta, file -> [ meta, "fusioncatcher", file ] }
+                FUSIONCATCHER_TO_CFF.out.cff
+                    .map{ meta, file -> [meta, file]}
             ).mix(
-                STARFUSION.out.abridged
-                    .map{ meta, file -> [ meta, "starfusion", file ] }
-            )
-            .groupTuple(by:[0],size:numcallers)
-            .map{ meta, caller, file ->
-                def avg_weight = caller.collect({(100/caller.size()).toInteger()})
-                avg_weight[-1] = avg_weight[-1] + (100-avg_weight.sum())
-                [ meta, caller, avg_weight, file ]
-            },
-        fusion_report_db
+                STARFUSION_TO_CFF.out.cff
+                    .map{ meta, file -> [meta, file]}
+            ).groupTuple(by:[0],size:numcallers),
     )
-    ch_versions = ch_versions.mix(FUSIONREPORT.out.versions.first())
-
-    CSV_TO_TSV(
-        FUSIONREPORT.out.fusionreport_csv
-            .map{ meta, csv ->
-                [meta, [csv]]
-            },
-        "csv",
-        "tsv"
+
+    METAFUSION(
+        MERGE_CFF.out.file_out,
+        gene_bed,
+        gene_info,
+        fasta,
+        blocklist
+    )
+
+    ADD_FLAG(
+        METAFUSION.out.cluster
+            .join(METAFUSION.out.cis, by:0)
+            .join(METAFUSION.out.cff, by:0)
+            .join(METAFUSION.out.problem_chrom, by:0)
+            .join(METAFUSION.out.filters, by:0)
     )
-    ch_versions = ch_versions.mix(CSV_TO_TSV.out.versions.first())
 
-    ONCOKB_FUSIONANNOTATOR(CSV_TO_TSV.out.csv)
+    ONCOKB_FUSIONANNOTATOR(ADD_FLAG.out.unfiltered_cff)
     ch_versions = ch_versions.mix(ONCOKB_FUSIONANNOTATOR.out.versions.first())
 
+    AGFUSION_BATCH(
+        ADD_FLAG.out.unfiltered_cff,
+        agfusion_db,
+        pyensembl_cache
+    )
+    ch_versions = ch_versions.mix(AGFUSION_BATCH.out.versions.first())
+
+    CFF_FINALIZE(
+        ADD_FLAG.out.unfiltered_cff
+            .join(ONCOKB_FUSIONANNOTATOR.out.oncokb_fusions, by:0)
+            .join(AGFUSION_BATCH.out.fusion_transcripts_tsv, by:0)
+    )
+    ch_versions = ch_versions.mix(ADD_FLAG.out.versions.first())
+    ch_versions = ch_versions.mix(METAFUSION.out.versions.first())
+    ch_versions = ch_versions.mix(ARRIBA_TO_CFF.out.versions.first())
+    ch_versions = ch_versions.mix(FUSIONCATCHER_TO_CFF.out.versions.first())
+    ch_versions = ch_versions.mix(STARFUSION_TO_CFF.out.versions.first())
+
     emit:
     ch_versions
 }
diff --git a/subworkflows/local/prepare_references.nf b/subworkflows/local/prepare_references.nf
index a45f70b..463daf6 100644
--- a/subworkflows/local/prepare_references.nf
+++ b/subworkflows/local/prepare_references.nf
@@ -5,11 +5,16 @@ include { GATK4_CREATESEQUENCEDICTIONARY } from '../../modules/nf-core/gatk4/cre
 include { SAMTOOLS_FAIDX                 } from '../../modules/nf-core/samtools/faidx/main'
 include { GATK4_BEDTOINTERVALLIST        } from '../../modules/nf-core/gatk4/bedtointervallist/main'
 include { PREPARE_RRNA                   } from '../../modules/local/prepare_rrna/main'
-include { GUNZIP                         } from '../../modules/nf-core/gunzip/main'
+include {
+    GUNZIP as GUNZIP_GTF ;
+    GUNZIP as GUNZIP_METAFUSIONGENEBED ;
+    GUNZIP as GUNZIP_METAFUSIONBLOCKLIST
+} from '../../modules/nf-core/gunzip/main'
 include { STARFUSION_DOWNLOAD            } from '../../modules/local/starfusion/download/main'
 include { FUSIONCATCHER_DOWNLOAD         } from '../../modules/local/fusioncatcher/download/main'
 include { FUSIONREPORT_DOWNLOAD          } from '../../modules/local/fusionreport/download/main'
 include { KALLISTO_INDEX                 } from '../../modules/nf-core/kallisto/index/main'
+include { AGFUSION_DOWNLOAD              } from '../../modules/local/agfusion/download/main'
 
 
 workflow PREPARE_REFERENCES {
@@ -18,12 +23,26 @@ workflow PREPARE_REFERENCES {
     ch_versions = Channel.empty()
 
     if (params.gtf.endsWith(".gz")){
-        GUNZIP([[:],params.gtf])
-        gtf = GUNZIP.out.gunzip.map{ it[1] }.first()
+        GUNZIP_GTF([[:],params.gtf])
+        gtf = GUNZIP_GTF.out.gunzip.map{ it[1] }.first()
     } else {
         gtf = params.gtf
     }
 
+    if (params.metafusion_blocklist.endsWith(".gz")){
+        GUNZIP_METAFUSIONBLOCKLIST([[:],params.metafusion_blocklist])
+        metafusion_blocklist = GUNZIP_METAFUSIONBLOCKLIST.out.gunzip.map{ it[1] }.first()
+    } else {
+        metafusion_blocklist = params.metafusion_blocklist
+    }
+
+    if (params.metafusion_gene_bed.endsWith(".gz")){
+        GUNZIP_METAFUSIONGENEBED([[:],params.metafusion_gene_bed])
+        metafusion_gene_bed = GUNZIP_METAFUSIONGENEBED.out.gunzip.map{ it[1] }.first()
+    } else {
+        metafusion_gene_bed = params.metafusion_gene_bed
+    }
+
     STAR_GENOMEGENERATE(params.fasta,gtf)
     ch_versions = ch_versions.mix(STAR_GENOMEGENERATE.out.versions)
     star_index = STAR_GENOMEGENERATE.out.index
@@ -72,6 +91,12 @@ workflow PREPARE_REFERENCES {
     //cosmic_passwd = params.cosmic_passwd ?: ""
     FUSIONREPORT_DOWNLOAD()
 
+    AGFUSION_DOWNLOAD(
+        params.ensembl_version,
+        params.genome
+    )
+    ch_versions = ch_versions.mix(AGFUSION_DOWNLOAD.out.versions)
+
     KALLISTO_INDEX(params.cdna)
     ch_versions = ch_versions.mix(KALLISTO_INDEX.out.versions)
 
@@ -90,6 +115,10 @@ workflow PREPARE_REFERENCES {
     fusion_report_db   = FUSIONREPORT_DOWNLOAD.out.reference
     rseqc_bed          = UCSC_GENEPREDTOBED.out.bed.map{it[1]}.first()
     kallisto_index     = KALLISTO_INDEX.out.idx
+    agfusion_db        = AGFUSION_DOWNLOAD.out.agfusion_db
+    pyensembl_cache    = AGFUSION_DOWNLOAD.out.pyensembl_cache
+    metafusion_blocklist = metafusion_blocklist
+    metafusion_gene_bed = metafusion_gene_bed
     ch_versions        = ch_versions
 
 }
diff --git a/tests/nextflow.config b/tests/nextflow.config
index bab5342..fdd8833 100755
--- a/tests/nextflow.config
+++ b/tests/nextflow.config
@@ -11,6 +11,10 @@ process {
     cpus = 2
     memory = 3.GB
     time = 2.h
+
+    withName: '.*_TO_CFF|GUNZIP_METAFUSION.*|AGFUSION.*' {
+        ext.when = { ! (workflow.profile.toString().split(",").contains(["singularity"]) && workflow.profile.toString().split(",").contains(["test"])) }
+    }
 }
 
 if ("$PROFILE" == "singularity") {
diff --git a/workflows/forte.nf b/workflows/forte.nf
index 23e9623..2569e4f 100644
--- a/workflows/forte.nf
+++ b/workflows/forte.nf
@@ -10,7 +10,7 @@ def summary_params = NfcoreSchema.paramsSummaryMap(workflow, params)
 WorkflowForte.initialise(params, log)
 
 // Check input path parameters to see if they exist
-def checkPathParamList = [ params.input, params.multiqc_config, params.fasta, params.gtf, params.refflat, params.starfusion_url ]
+def checkPathParamList = [ params.input, params.multiqc_config, params.fasta, params.gtf, params.refflat, params.starfusion_url, params.metafusion_gene_bed, params.metafusion_blocklist, params.metafusion_gene_info ]
 for (param in checkPathParamList) { if (param) { file(param, checkIfExists: true) } }
 
 // Check mandatory parameters
@@ -130,7 +130,11 @@ workflow FORTE {
         PREPARE_REFERENCES.out.gtf,
         PREPARE_REFERENCES.out.starfusion_ref,
         PREPARE_REFERENCES.out.fusioncatcher_ref,
-        PREPARE_REFERENCES.out.fusion_report_db
+        PREPARE_REFERENCES.out.fusion_report_db,
+        PREPARE_REFERENCES.out.agfusion_db,
+        PREPARE_REFERENCES.out.pyensembl_cache,
+        PREPARE_REFERENCES.out.metafusion_gene_bed,
+        PREPARE_REFERENCES.out.metafusion_blocklist
     )
     ch_versions = ch_versions.mix(FUSION.out.ch_versions)