2013-01-29

Pipeline 5.0.2

Count mapped reads which have > 100bp alignment.

2014-01-30

pipeline 5.0.3

--local (required) - end trimming

-M 100 (remove)

-N 0 (required) - 0 mismatches to seed

-L 20 (modify) - specifies seed length - should be set to 1/2 minimum alignment length since we're requiring 100% identity

-i S,1,0 (modify) - need to reduce the number of seeds tested - to test every 5 seeds from a 300mer, the combo would be S,1,0.23

--mp 1000,1000 (reqired) - all scoring options seem OK for a minimum alignment length of 50

--np 1000

--rdg 1000,1000

--rfg 1000,1000

--score-min L,100,0

--no-mixed (remove)

--fr (required)

--no-discordant (required)

--ignore-quals (required)

count reads with alignment > 50bp

2013-01-22

Change to pipeline 5.0.1

1. remove the 30bp exon-intron boundary.

DESeq

1. input csv file has CR will ruin everything
2. GLM modeling, need to discuss with Julia

2014-01-23

Modification in 5.0.1

countMappedReads.py

1. Remove the 30bp intron-exon boundary condition in calculating the coverage
2. Generate max coverage file to record the max overall reads coverage in each exon
3.  Do not generate filter reads any more.
4. Sort all individual reference with max overall reads coverage, then pickup the top 200.

HLA_parts.sh

1. Do not remove the scratch directory

map_single.sh

1. Do not remove the scratch directory

select_single.sh

 1. Remove filtered reads location as input parameters

runTyping.py

1. use /BioScratch/zfu as scratch dir
2. no filtered reads parameters
3. use one countMappedReads.py for all classes

/Bioinformatics/Users/zfu/HLA_Typing/HLA_cDNA_Database/IMGT.Release.3.12.0/Index.Test.Single

Only kept alleles with 2-digits subtype name

work log 01-21-2014

1.  Finished the recommendation letter 3
2. Generate reports for HLA Run36