2014-04-24

update the model file and train feature file in

10.0.3.183/srv/www/classifier_tool/classifier_files/SVM_2012/

update the weka model file in

10.0.3.183/srv/www/classifier_tool/classifier_files/Weka_model_files/

wrong ids in curatable and uncuratable list due to the "\n" by using "cat" to combine files

2014-04-18

7.0.1

Class I + DRB1

Search for allele which has one read cover the entire exon for each exon (2, 3,4)

Class II

Search for allele has coverage on each base at exon 2 + exon 3

7.0.2

Exclude 10 bases from exon-intron boundary and find the minimum coverage

7.0.3

find the minimum coverage from the exon-intron boundary

Search for allele which has one read cover the entire exon for each exon (2, 3,4) , if its number < 2, then search alleles have coverage on each base at exon 2 + exon 3 + exon 4

2014-04-02

find . -maxdepth 6 -type f -name "depthOfCoverage_allReads.txt" -exec cp --parents {} /Bioinformatics/Users/zfu/HLA_Typing/Run42_Coverage \;

find . -maxdepth 6 -type f -name "depthOfCoverage_allReads.txt" -exec cp --parents {} /Bioinformatics/Users/zfu/HLA_Typing/Run43_Coverage \;

2014-04-01

1. add a count for the number of mapped reads to each exon in the output
2. add a count for the total bases mapped in each exon to the output
3. use the count of the total bases mapped to the templates to select the best pairs rather than the total number of reads mapped
4. enforce continuous coverage within each exon by adding several constraints
5. reads that cross exon borders should only be counted in the exon in which the majority of the read lies

dict 1:

key: allele id
value: dict 2

dict 2:

key: exon
value: dict 3

dict 3:

key: short read id + sequence
value: list 4

list 4: alignment start , alignment length, alignment end