ComboRATE:
Iteratively group HLAs to lower RATE p-value
Install the dependencies with
pip install -r requirements.txt
USAGE:
python comborate.py -a <allele file> -r <response file> -p <IEDB prediction file> -o <output directory>
options: -h, --help show this help message and exit -a, --allele-file ALLELE_FILE Allele file name -r, --response-file RESPONSE_FILE Response file name -o, --output OUTPUT Output dir name
Optional: -p, --prediction PREDICTION HLA prediction file name -c, --rank-cutoff RANK_CUTOFF Cutoff for the binding rank (default = 25) -k, --response-cutoff RESPONSE_CUTOFF Cutoff for the response (default = 1)
Run test with
python comborate.py -a test/allele.txt -r test/response.txt -p test/iedb_netmhciipan_el_result.csv -o test_result
The test results will be in the test_result folder created by the program in the test above and contain, for each peptide, two files:
<peptide>_pred.csv with the IEDB predictions isolated for the peptide
<peptide>_rate.csv with the RATE results for the peptide
In the _rate.csv file, you will also find the ComboRATE groups of restrictions, with the number of the iteration in which the group was created and the alleles that compose the group. E.g.:
The 5 first rows test_result/01_rate.csv read, for the first column:
2-DQA1*03:01/DQB1*03:02+DRB1*04:07+DRB1*07:01
1-DQA1*03:01/DQB1*03:02+DRB1*04:07
DQA1*03:01/DQB1*03:02
DRB1*04:07
DRB1*07:01
Lines 3 and 4 contain the restrictions with the two best p-values DQA1*03:01/DQB1*03:02, DRB1*04:07, which were grouped in the first iteration, making the group 1-DQA1*03:01/DQB1*03:02+DRB1*04:07, which means
1- (first iteration)
DQA1*03:01/DQB1*03:02 (restriction with the lowest p-value)
DRB1*04:07 (restriction with the second lowest p-value)
Later, ComboRATE added DRB1*07:01 to the 1st group, making 2-DQA1*03:01/DQB1*03:02+DRB1*04:07+DRB1*07:01, which means
2- (second iteration)
DQA1*03:01/DQB1*03:02 (restriction with the lowest p-value)
DRB1*04:07 (restriction with the second lowest p-value)
DRB1*07:01 (restriction with the third lowest p-value)