pseudocode_deduper

DEDUPER PSEUDOCODE

LOCATION
    /home/ano/bioinfo/Bi624/Deduper-Dremtz

INPUT
    SAM file

OUTPUT: all SAM files
    if computational time
    1. deduplicated
    2. duplicates that were removed from output1
    3. umatched/error UMIs
GOAL
    Remove all PCR duplicates from SAM file leaving only copy remaining in the SAM file
    
SAM headers
RNAME (col 3) -> chromosome
POS (col 4) -> position
FLAG (col 2) -> strandedness
CIGAR (col 6) -> soft clipping
QNAME (col 1) -> UMI

PIPELINE OVERVIEW


After reading in files, UMI list and opening a loop

BASH
sort by position start. ###I think this will correctly sort reverse strands also but check (use samtools for this?)

PYTHON
###Dre, is this the most efficient order?

    1. IF Sequence matches or is a close match to an UMI (QNAME field) then don't return,    
    DO return WANTED sequences

    2 IF clipping happened (CIGAR str in bitwise flag #)
        THEN ignore clipped sections and compare Position and Sequence from after clipped section    ###Dre: How do I do this?

    3. IF Chromosome Name (SAM field 3, CHROMNAME) 
                AND IF forward strand (SAM field 2)
                        AND left most Position matches (SAM field 4, POS)
                            AND sequence MATCHES
                                THEN write a single sequence to new file

                        OR reverse stranded (SAM field 4, POS)
                            AND right most Position matches ###DRE how do I check for this?
                                AND sequence MATCHES
                                    THEN write a single sequence to new file

    4. RETURN list that does not have any of the excluded sequences from above

FUNCTIONS:

Function 1: maybe write a function to grab just the fields we want from SAM
    put these in to a dictionary with key as UMI

Function 2: adjust position based off of clipping number

Function 3: if reverse compliment, then account for indels
maybe make this an addon to function 2

UNIT TESTS:
    /home/ano/bioinfo/Bi624/Deduper-Dremtz/unit_tests
    #delete in unit test descriptions before running UT's

    EXPECTED OUTPUT = 7 Lines
    2 unmatched chromname same everything else
    1 clipped file that matches another file but is clipped 2 spaces
    1 matching fwd strand w/ same starting Positions
    2 fwd compliment with different starting positions
    1 rev compliments w/ same ending position and sequence
    = 7 total lines