README for POTION
             (last updated 12/02/2015 - mm/dd/yyyy format)

AUTHORS
-=-=-=-

Jorge Augusto Hongo ([email protected])
Giovanni Marques de Castro ([email protected])
Francisco Pereira Lobo ([email protected], [email protected])


1 - DESCRIPTION
-=-=-=-=-=-=-=-

POTION is a software to find groups of homologous genes under positive
selection. Our pipeline links several third-party software to perform
the steps needed for positive selection detection. For this reason, you
must install these software to use POTION. See the INSTALL file for
further instructions about them.



2 - HOW TO USE - IN BRIEF 
-=-=-=-=-=-=-=-=-=-=-=-=-

2.1 - INSTALL 
-=-=-=-=-=-=-

Download POTION's dependencies detailed in the INSTALL file and set the
parameters in 'potion_config' file ('config_files' directory).

POTION comes with two examples (MYC and TRYP datasets, 'examples' folder) that
we used to validate it. We recommend you to test your installation with the
TRYP dataset (requires MUSCLE and dnaml) due to the long running time expected
for the MYC dataset.

To test POTION with the TRYP dataset, go to the bin/ folder and type the
following:

  perl potion.pl --conf_file ../examples/TRYP/config_models/TRYP.conf

If you had no error messages, check the 'final_results.out' file in
'POTION-<version>/results_TRYP/'. It should have around 41 lines (including
header) detailing the evidence of positive selection in the Trypanosoma brucei.
  

2.2 - PREPARING INPUT FILES 
-=-=-=-=-=-=-=-=-=-=-=-=-=-

Once installed, there are three steps to use POTION:

A) Generate the following input files:
    - Fasta files with the nucleotide coding DNA sequence (CDS) of all genes of
      each species (one species per file). Keep them in a single directory.
    - OrthoMCL v1.4 output file with gene homology relationships.
    
You can use the scripts from 'bin/utils' directory to generate the fasta and
orthomcl v1.4 files from genbank/orthoxml formats.

B) Create your project's configuration file (user_config_file models). 

To create a default configuration file, execute POTION with '--create_conf'
option. POTION fills most parameters if you provide the proper information
during the process. The default parameters left to fill are:
  1) Path to the fasta files  ('CDS_dir_path')
  2) OrthoMCL 1.4 output file ('homology_file_path')
  3) Output directory's name  ('project_dir_path')
    
For a complete list of configuration parameters, execute POTION with the 
'--help' option.
    

2.3 - RUNNING POTION 
-=-=-=-=-=-=-=-=-=-=-    

Execute POTION, using as input the project's configuration file from step B:

  perl potion.pl --conf_file <path to your project's configuration file>



3 - HOW TO USE - IN DETAIL 
-=-=-=-=-=-=-=-=-=-=-=-=-=

3.1 - INSTALL 
-=-=-=-=-=-=-

Install POTION following the INSTALL file instructions. Check if everything
went right by executing POTION with the TRYP dataset ('examples' folder) by
typing the following from the bin/ folder:

  perl potion.pl --conf_file ../examples/TRYP/config_models/TRYP.conf

The install procedures will create a main directory called POTION-<version>.
This directory contains the following subdirectories:

  bin - contains POTION's main files (potion.pl and module_latest.pl), as well
as a subdirectory ('utils') with scripts to assist the generation of the input
files for POTION.

  config_files - contains empty configuration files that POTION uses to work
with third-party software, to create a default project configuration file
(option '--create_conf') and the 'potion_config' file, which you must edit for
POTION to find the third-party software in your machine.
  
  examples - contains the two example datasets (TRYP and MYC) used to validate
POTION. Use them to validate your POTION installation. For each dataset, we
provide the homology relationship data, sequence data and POTION configuration
files to reproduce the results described in our article. 


3.2 - RUNNING POTION WITH EXAMPLE DATA 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

We distribute POTION with two example datasets (TRYP and MYC) that can be used
to reproduce our results ('examples' directory). To execute POTION using these
datasets, edit the configuration file and point an output directory. Please
notice that POTION produces a considerable amount of intermediate files. A
complete analysis of the MYC dataset, for instance, can easily create results 
of more than 5 Gb.


3.3 - POTION INPUT FILES -
=-=-=-=-=-=-=-=-=-=-=-=

You need to provide a multi-fasta CDS file for each species and a file
describing the homology relationships among CDS. Sequence IDs in fasta
files must be the same found in the file containing homology data.

The file describing homology relationships is identical to OrthoMCL 1.4
output, which describes one group of homologous per line as follows:

GROUP1(2 genes,2 taxa):	gene_name_A(genome_1.fa) gene_name_B(genome_2.fa)

The colon and the count of genes/taxa are optional. A tab character
("\t") is obligatory to separate the homologous group name from its
genes/specie. Each gene is described by two pieces of information:
gene_name (must be the same ID used in fasta files) and genome_name
(between parenthesis, must be the same name of fasta file where gene
is located).

This format assumes no use of spaces or parenthesis when naming genes or
species. Avoid using these characteres when naming your files and
genes, as they can generate ambiguous interpretations for this format.


3.4 - PREPARING YOUR OWN DATA 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

POTION is distributed with several scripts to automate data acquisition
('bin/utils/' directory) and allow users to analyze their own data.
These scripts can automatically download genbank text files and extract
CDS and protein data. More details about these scripts is found in
"README_utils.txt", also located in the 'utils' directory.

To obtain homology data users can: 1) Infer protein data with OrthoMCL
v1.4 and use its main output file or 2) Parse OrthoXML data to the
OrthoMCL output format. 

The 'orthoXML2MCL.pl' script is offered to use data from specialized
databases of homologous genes that export homology information in
OrthoXML format.

POTION's behavior is dictated through a project configuration file. To
create an empty project configuration file, just execute POTION with the
'--create_conf' option.


3.5 - SETTING UP POTION PARAMETERS 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

POTION contains two configuration files. One of them (install configuration
file) is used to configure POTION parameters that are not expected to change
after POTION install, such as the location of third-party programs, POTION's
root install directory, and the name of output files. This configuration file
is meant to be specific to each machine.

The other configuration file (project configuration file) contains parameters
that are specific to each project, such as location of sequence files, filters
to be applied to data, or cutoffs to remove sequence and groups. This
configuration file is meant to to allow you to share your project with other
users, so that they can reproduce your results with your data. Our MYC and TRYP
datasets, in the 'examples' folder, are an example of that.

Variables that point to directory paths should end with "/" whenever possible.
Comment lines in these files are ignored. A special variable ($potion_dir) is
defined in the install configuration file, and should point to the complete
path to POTION's folder. The variable 'potion_dir' can be used in the
parameters of both configuration files (install and project) as a shortcut to
POTION's directory. For example, if you have:

  potion_dir = /home/myname/POTION-<version>/

and want to store your results in a directory named "analysis_yyyymmdd",
then both lines refer to the same directory:

  project_dir = /home/myname/POTION-<version>/projects/analysis_yyyymmdd/
  project_dir = $potion_dir/projects/analysis_yyyymmdd/

  
  

3.5.1 - POTION install configuration file 
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

This file is located in directory 'config_files/' as 'potion_config'.

-------------- + potion_dir + --------------

potion_dir - path to the root directory of POTION.
             You can use '$potion_dir' in other parameters as a shortcut
             to the path you defined in this parameter.

------------------- + paths to third-party programs + ---------------

codeml - path to the Codeml executable of PAML package 
consense - path to the Consense executable of Phylip package 
dnaml - path to the Dnaml executable 
muscle - path to the MUSCLE executable 
phipack - path to the Phi executable of PhiPack 
prank - path to the Prank executable 
proml - path to the Proml executable of Phylip package 
seqboot - path to the Seqboot executable of Phylip package 
trimal - path to the Trimal executable 
mafft - path to mafft executable 
phyml - path to phyml executable

-------------- + names of output files + -------------------------

result_table - name of main table file. [Default: final_results.out]

result_uncertain - name of fasta files containing sequences with some
evidence of positive selection, but not strong enough after FDR.
[Default: uncertain.out]

result_positive - name of files containing evidence of positive
selection after FDR. [Default: positive.out]

result_recombinants - name of files containing sequences with evidence
of recombination. [Default: recombinants.out]

summary - name of files containing summary of POTION execution.
[Default: summary.out]

tries - number of tries for each program, in case it crashes or returns
an error.


3.5.2 - POTION project configuration file
-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

POTION contains a rich set of parameters to dictate how it should behave
during execution. Some of them are project-related, such as to define
input and output files and directories, or the number of processors to
be used in the parallelized steps of POTION. Other parameters are the
several filters and customizations to taylor POTION to user-specific
needs, such as to analyze only putative 1-1 orthologs or paralogs, to
remove sequence outliers in terms of quality, length or identity.
Finally, at several steps POTION allow users to choose between distinct
third-party software, or to configure some steps of these software. 

Below follows the explanations of the parameters currently available in
POTION. An "M" indicates a mandatory parameter.


PROJECT PARAMETERS 
-=-=-=-=-=-=-=-=-=

mode - POTION's main analysis mode. Currently POTION only supports
"site" as main analysis mode, future versions are expected to support
"branch" and "branch-site" modes as well.

CDS_dir_path - path to the directory containing the fasta CDS files. (M)

homology_file_path - path to the OrthoMCL 1.4 format file, describes the
homology relationships among the CDS in the fasta files. (M)

project_dir_path - path to the directory where the results will be written.
If the directory doesn't exist, POTION will attempt to create it. (M)

max_processes - Defines the maximum number of processors to be used in
the parallelized steps of POTION's algorithm. (M)

remove_identical - Defines if POTION should remove groups 100% identical
at nucleotide level. Possible values are "yes" and "no". (M)

verbose - 1 to print nice log messages telling you what is going on, 0
otherwise. (M)


SEQUENCE/GROUP PARAMETERS 
-=-=-=-=-=-=-=-=-=-=-=-=-

groups_to_process - Defines which lines of the clusters file will be
processed. Use "all" to analyze every group (default), "-" to set
groups between two given lines (including the said lines). Use "!" to
not process a specific line, can be used with "-" to specify a set to
not be processed. Use "," or ";" to set distinct line sets. Examples:
  "1;4-10;12" will process groups described in lines 1, 4 to 10, and 12.
  "all;!3"    will process all groups, except group in line 3.
  "all;!3-5"  will process all groups, except groups in lines 3 to 5.

behavior_about_bad_clusters - Defines what POTION should do if it finds
a group of homologous that contain any gene removed by a filter.
Possible options are:
  "0" (does not filter any sequence, not recommended).
  "1" (remove only sequences and keep group).
  "2" (remove all groups containning flagged sequences).
Sequences can be flagged by any of the following parameters below: 
validation_criteria, absolute_min_sequence_size, absolute_max_sequence_size,
relative_min_sequence_size, relative_max_sequence_size,
identity_most_similar_sequence, mean_sequence_identity. (M)

behavior_about_paralogs - Defines how POTION should handle mixed groups
of homologous genes that contain orthologs and/or paralogs. Possible
values are:
  "0" analyze all sequences within group.
  "1" remove paralogs and analyze only the remaining putative 1-1 orthologs.
  "2" remove groups with paralogous.
  "3" remove 1-1 orthologs and analyze all remaining paralogs together.
  "4" remove 1-1 orthologs and split remaining paralogs to analyze each species 
      individually. (M)

validation_criteria - Defines which quality criteria POTION should use to 
remove individual sequences. Possible values are: 
  "0"   do not perform any sequence quality filter.
  "1"   checks for valid start codons.
  "2"   checks for valid stop codons.
  "3"   checks for sequence length multiple of three.
  "4"   checks for non standard nucleotides.
  "all" applies all previous criteria.

additional_start_codons - these codons, plus the ones specified in codon table,
will be the valid start codons for validation purposes. "(CTG,GTG)", for
instance, will define these two codons as additional start codons. Parenthesis
and commas to enumerate more than one codon are mandatory.

additional_stop_codons: Same as above, for stop codons.

codon_table - Indicates the codon table id (1 to 22) to be used by POTION for
translation and validation purposes, possible tables are defined in NCBI and
bioperl. See http://www.ncbi.nlm.nih.gov/Taxonomy/Utils/wprintgc.cgi for the
definition and taxonomic range of the 22 official genetic codes. (M)

absolute_min_sequence_size - minimum absolute sequence length cutoff.
Sequences whose lengths are smaller than this value are removed from
further analyses. Possible values are all non-negative integers (ex: 200).

absolute_max_sequence_size - maximum absolute sequence length cutoff.
Sequences whose lengths are greater than this value are removed from
further analyses. Possible values are all non-negative integers (ex: 400).

relative_min_sequence_size - minimum relative sequence length cutoff.
Sequences smaller than mean/median times this value will be removed.
Possible values are numbers within the interval of 0 and 1. Mean/median
values are calculated for the whole group in order to allow the comparison.

relative_max_sequence_size - maximum relative sequence length cutoff.
Sequences greater than mean/median times this value will be removed.
Possible values are numbers between 1 and infinite. Mean/median values
are calculated for the whole group in order to allow the comparison.

sequence_size_average_metric - Dictates which metric will be calculated
to determine the minimum/maximum relative lengths ranges for sequence removal
(configured in [relative_min|relative_max]_sequence_size). Possible values are
"mean" or "median". (M if relative_min or relative_max are defined).

min_group_identity - mean minimum group identity cutoff in pairwise
sequence alignments.

max_group_identity - mean maximum group identity cutoff in pairwise
sequence alignments.

group_identity_comparison - indicates the kind of sequence that will be
used when computing mean group identity. Possible values are "nt" and "aa".

min_sequence_identity - minimum (mean/median) sequence identity cutoff
in pairwise sequence alignments.

max_sequence_identity - maximum (mean/median) sequence identity cutoff
in pairwise sequence alignemnts.

sequence_identity_average_metric - dictates which metric will be calculated to
identify sequences with extreme values of identity (configured in 
[min|max]_sequence_identity). Possible values are "mean" and "median".

sequence_identity_comparison - indicates the kind of sequence that will
be used when computing sequence identity. Possible values are "nt" and "aa".

minimum_gene_number_per_cluster - lower bound cutoff for the minimum number of
genes in group after all filtering steps.

maximum_gene_number_per_cluster - upper bound cutoff for the maximum number of
genes in group after all filtering steps.

minimum_specie_number_per_cluster - lower bound cutoff for the minimum number
of species in group after all filtering steps.

maximum_specie_number_per_cluster - upper bound cutoff for the maximum number
of species in group after all filtering steps.

reference_genome_files - This is the final filter step applied by POTION. If
users chose a reference genome file, POTION will only analyze groups that,
after all previous filtering steps, contain at least one gene from reference
genome. POTION will also return all the results found (sequence, IDs and
coordinates) in reference to the anchor genome. Possible value is the name of
one fasta CDS file.


THIRD-PARTY SOFTWARE CONFIGURATION 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-

multiple_alignment - defines program used for multiple protein sequence
alignment. Possible values are muscle|mafft|prank. Muscle is executed with
default parameters, prank is executed with the flags -twice (to run the
analysis twice for each group) and -F (to correctly penalize the gaps) and
mafft is executed with the flag --auto for auto-configuration. (M)

bootstrap - Number of bootstraps to be generated in the phylogenetic analysis.
Possible values are all non-negative integers. (M)

phylogenetic_tree - POTION currently supports proml|dnaml from phylip
package and phyml. All of them calculate phylogenetic trees using the
maximum likelihood approach on protein or nucleotide data. Possible
values are proml|dnaml|phyml_nt|phyml_aa. (M)

phylogenetic_tree_speed - Both proml and dnaml contain a fast execution
mode. Use this parameter to turn it on or off. Possible values are
fast|slow. (M)

rec_minimum_confirmations - This number dictates the minimum number of
recombination tests with significant q-values (FDR-corrected p-values)
in order to infer recombination in a group. Possible values are 0, 1, 2
and 3. A value of 2, for instance, will demand significance in at least
two recombination tests to infer recombination occurrence.

rec_mandatory_tests - Dictates which recombination tests must be
positive in order to infer recombination. Possible values are "phi",
"maxchi2" and "nss".
It is possible to request more than one test. For instance, setting this
variable as "phi maxchi2" will request POTION to find recombination for both 
phi and maxchi2 tests to classify a group as with evidence of recombination. 
The configuration of this variable can be used with rec_minimum_confirmations.
Setting these variables for "phi" and "2", for instance, will request
recombination detection in at least two tests, and one of them must be phi. The
definition of the three recombination tests can be found in PhiPack package.

recombination_qvalue - Defines the q-value for recombination detection.
It must occur for all the specified tests. Possible values are numbers
between 0 and 1. If 0, POTION will not look for recombination.

remove_gaps - This variable controls how POTION should trim columns of
multiple sequence alignments:
  - If filled with numeric values between 0 and 1, POTION will use this number
    as upper bound cutoff for the number of gaps in a given column alignment. 
  - The values of "strict" or "strictplus" for this variable will execute more
    complex column filters that take into account the conservation of the
    neighborhood of a given column to trim sequence alignments. These filters
    are described in detail in the supplementary material of the article that
    describes trimAl, the software that performs this step.

PAML_models - This variable configures which nested codon evolution
models will be computed by codeml. Possible values are "m12", "m78"
and/or "m8a8". To evaluate all nested models, write "m12 m78 m8a8" as
value for this variable. (M)

pvalue - p-values for positive selection detection. Groups with p-values
lower than this cutoff and greater q-value cutoff are marked as "u"
(uncertain) in POTION's main result table and in the respective fasta
files (please check the respective sections below to obtain further
information about POTION output files).

qvalue - q-values for positive selection detection. Groups with q-values
lower than this cutoff are marked as "P" (positively-selected) in POTION's
main results table and in the respective fasta files (please check the
respective sections below to obtain further information about POTION output
files).


3.6 - POTION OUTPUT 
-=-=-=-=-=-=-=-=-=-

POTION produces detailed intermediate and final results for each group.
All final results are stored in 'results' directory, intermediate files
are stored in 'intermediate_files' directory.

Our software also produces detailed log information about each sequence
and each group ('log' and 'log.err' files). POTION also creates a copy
of the project configuration file in your project's output directory.


MAIN OUTPUT FILES 
=-=-=-=-=-=-=-=-=

Currently, our software returns the following, whose names can be
configured in 'potion_config' file ('config_files' directory):


-------------------- + final_results.out + --------------------

Tabular flat file containing the statistical data and final classification
for each valid group (i. e. not removed by a filter). The values in this
file are:

 - cluster name.

 - final classification for nested models M1a2 and M78:
    - "P" for groups with q-values smaller than the cutoff.
    - "u" for groups with q-values greater than the cutoff and p-values smaller
      than the cutoff.
    - "n" for negative groups.
    - "-" for groups where it was not possible to compute p-values.

 - -lnL values for models evaluated (M1a, M2, M7, M8a, M8). Possible
   values are a negative number, and "-" for groups where it was not
   possible to compute a -lnL value.

 - D values, the difference of likelihoods for neutral models (M1a, M7
   and M8a) when compared with models that allow sites under positive
   selection (M2, M8 and M8, respectively).

 - p-values computed for each pair of nested models evaluated.
   Calculated through a Likelihood Ratio Test (LRT).

 - q-values computed for each pair of nested models evaluated.
   Calculated through a False Discovery Rate correction.


---------------- + allvalid_for_enrichment + ----------------------

Fasta flat file containing all valid groups after filtering steps, but
prior to recombination detection. Meant as a background distribution for
enrichment analysis if you want to compare positively selected genes
against the background of valid genes, without considering recombination.


---------------- + non_recombinants_for_enrichment + -----------------

Fasta flat file containing all valid groups after filtering steps after
recombination detection. Meant as a background distribution for enrichment 
analysis if you want to compare positively selected genes against the
background of valid genes with no evidence of recombination.


---------------- + positive.out + ----------------

Flat file that contains information about sites under positive selection
found in each group of homologous genes. Each group is described in four
lines:

First line:
<Group ID>: <gene ID 1> <gene ID 2> ... <gene ID n>

Second line:
><gene id> (model where positive selection was detected)

Third line: sequence

Fourth line: sites under positive selection, where "-" indicates no
evidence, "+" indicates Bayes Empyrical Bayes (BEB) values greater than
0.95, "*" indicates BEB values greater than 0.99.


------------------- + positive.out.interleaved +-----------------------

Same as positive.out, but displayed in interleaved format to make it
easier to visualize the sites under positive selection.


------------ + positive.out.[aa|nt].*_for_enrichment + -----------------

Fasta flat files containing protein or nucleotide sequence data for
groups of homologous found as positively selected (q-values smaller than
cutoff). POTION will create at most three distinct files to store genes
found as positively selected for m2, m8 and both pairs of nested models.


------------ + recombinants.[aa|nt].*_for_enrichment + -----------------

Fasta flat files containing protein or nucleotide sequence data for
groups of homologous with evidence of recombination (q-values smaller
than cutoff).


--------------- + summary.out + ---------------

Flat file containing information about POTION execution, such as total
number of valid clusters, time spent, theoretical serialized time spent
and total time spent in each of the parallelized steps of POTION.


--------------- + uncertain.* + ---------------

Flat files containing similar information as the one produced for positively
selected genes. Same naming conventions.



INTERMEDIATE FILES 
=-=-=-=-=-=-=-=-=-

The directory 'intermediate_files' contain a directory for each valid
group submitted to the parallelized portion of POTION (starting with
multiple protein sequence alignment). Inside each group directory are
all intermediate files created during POTION's execution. File contents
are as follows:


id_names - links ids to internal ids used by POTION. Needed since
several programs/formats have requests that may not be fulfilled by
initial gene ids such as phylip maximum sequence name length of ten.

VALIDGROUP.group_status - if exists, this file contains information
about the reasons why the computation of this group could not go trough
the entire POTION pipeline. It may be the error message of third-party
software, if POTION was able to capture it, or some group filtering
criteria.

VALIDGROUP.cluster.aa.fa - amino acid fasta sequence file 

VALIDGROUP.cluster.nt.fa - nucleotide fasta sequence file

VALIDGROUP.cluster.aa.fa.aln - aligned amino acid fasta sequence file

VALIDGROUP.cluster.aa.fa.aln.1.fas - intermediate alignment file

VALIDGROUP.cluster.aa.fa.aln.[aa|nt].phy - aligned sequence in phylip
format for phylogenetic analysis.

VALIDGROUP.cluster.aa.fa.aln.[aa|nt].phy.trim - aligned trimmed sequence
in phylip format for phylogenetic analysis.

VALIDGROUP.cluster.aa.fa.aln.[aa|nt].phy.trim.boot - bootstrap file in
phylip format for phylogenetic analysis

VALIDGROUP.cluster.aa.fa.aln.[aa|nt].phy.trim.boot.log - execution log
for bootstrap program, phylip package

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.conf.boot - configuration file
used to create bootstrap file using bootstrap program, phylip package

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.conf.final_tree - configuration
file used to create consensus tree using consensus program, phylip package

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.conf.tree - configuration file
used to create phylogenetic trees using program [dna|pro]ml, phylip package,
or PhyML

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.final_tree - consensus tree,
newick format, produced using consensus tree, phylip package

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.final_tree.log - execution log
of consensus program, phylip package

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.tree - trees produced using bootstrap
files, produced using [dna|pro]ml, phylip package or PhyML software

VALIDGROUP.cluster.aa.fa.aln.aa.phy.trim.tree.log - execution log of
[dna|pro]ml, phylip package or PhyML software

VALIDGROUP.cluster.aa.fa.aln.nt.fa - codon alignment data, produced
using protein alignemnt data

VALIDGROUP.cluster.aa.fa.aln.nt.trim.fa - codon alignment trimmed data,
produced using trimal

VALIDGROUP.cluster.aa.fa.aln.nt.phy - codon alignment data, phylip format

VALIDGROUP.cluster.aa.fa.aln.nt.phy.trim - trimmed codon alignment data,
phylip format

VALIDGROUP.cluster.aa.fa.aln.nt.phy.trim.paml.model[1|2|7|8] - result of
codeml execution for model M1a, M2, M7 or M8.

VALIDGROUP.cluster.aa.fa.aln.nt.phy.trim.paml.model[1|2|7|8].config -
configuration file used to execute codeml for model M1a, M2, M7 or M8.

VALIDGROUP.cluster.aa.fa.aln.nt.phy.trim.paml.model[1|2|7|8].log -
execution log of codeml for model M1a, M2, M7 or M8.

VALIDGROUP.cluster.aa.fa.aln.pairs - contains information about group
and sequence identity metrics. Calculated using trimal.

VALIDGROUP.trimal_cols.aa - contains columns positions filtered out by
trimal. Used to trim codon alignment data.

Phi, Phi.inf.list, Phi.inf.sites, Phi.poly.unambig.sites - result and
log files produced by phipack package when searching for recombination

2NG.dN, 2NG.dS, 2NG.t, 4fold.nuc, lnf, rst, rst1, rub - intermediate
files produced by codeml

time_f_tree, time_id_rec, time_model_1, time_model_2, time_model_7,
time_model_8 - time used to compute multiple sequence alignment and
phylogenetic tree, recombination, and codeml models 1, 2, 7 and 8


log files 
-=-=-=-=-

POTION generates an execution log where detailed information about each
group and sequence can be found, including the reasons for removal from
further analyses.


3.7 - FREQUENTLY ASKED QUESTIONS 
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

Q - How did you choose the parameters for the "soft", "standard" and
"hard" filters?

A - Intuition and a lot of manual inspection of alignment data! In
brief, "soft" may contain some false-positives, but is useful if you did
not find any result when using POTION in "standard" or "hard" mode and
wish to see if you find anything suspicious for further investigation.
"standard" appears to give good results and removes most of the noisy
data, and usually is our first choice when analysing data with POTION
for the first time. "hard" is just what it means to be: it will remove
almost all data with looks barely suspicious, but what remains is very
likely to represent true positive selection, at least in our current
knowledge of POTION's behavior.

Q - How POTION behaves if there is more than one gene from the anchor
genome in a given group?

A - POTION will report only the longest ORF if you are running any
analysis where more than one gene from the same species is to be
reported. Possible cases are analyses of paralogs or analyses anchored
to a given genome.


3.8 - DEVELOPMENT NOTES/KNOWN ISSUES
=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=

Some notes about the third-party programs/module:


- New versions of BioPerl could cause POTION to get stuck when transforming
alignment files from one format to another. Newer versions of POTION (from
1.1.2 on) do not rely on AlignIO BioPerl module for this step.

- Prank currently uses the parameters -twice, -F and -quiet; the second
file produced as output is used as the alignment

- Some programs can crash during execution even if they would run normally if
done manually. Programs that have any bug regarding memory leaks have a higher
likelihood to run into it when used in a highly-parallelized pipeline; POTION
will attempt to isolate the problem and re-run the program at least three times
(can be set by the user) before excluding the group from the analysis.

- Prank may abort a few times for a large number of processes (we are not sure,
but it appears to be due to memory issues; POTION will attempt to reinitialize
it until it runs properly. If rerunning doesn't solve it, we recommend running
with less processes, reducing the parameter 'max_processes' until the problem
stops occurring.

- If you want to halt prank during a run, use 'killall perl' followed by
'killall prank'. However, it doesn't prevent possible errors due to memory
overwrite.

- "version `GLIBC_x.xx' not found" is a possible message when running PAML,
Phylip executables and other programs written in C. This indicates that they
were compiled with a different version of C library than the one in the 
computer you are using. Update your C library (GLIBC) from 
http://www.gnu.org/software/libc/index.html or recompile the executable with
your current library.