search

TRON-Bioinformatics / seq2HLA

In-silico method written in Python and R to determine HLA genotypes of a sample. seq2HLA takes standard RNA-Seq sequence reads in fastq format as input, uses a bowtie index comprising all HLA alleles and outputs the most likely HLA class I and class II genotypes (in 4 digit resolution), a p-value for each call, and the expression of each class.
MIT License
47 stars 7 forks source link

error running seq2HLA on mRNAseq #18

Closed katahdin0 closed 1 month ago

katahdin0 commented 1 month ago

The error below seems to be arising from some software version incompatibility ? but it is not clear what. bowtie log is not empty, and bowtie version is 1.3.1

The digital haplotype is written into run-ClassI-class.digitalhaplotype2 Took 0.26239871978759766 seconds 2nd iteration done. -----------2 digit typing results-------------

Locus Allele 1 Confidence Allele 2 Confidence

A no NA hoz("A01") NA B no NA hoz("B07") NA C no NA hoz("C*01") NA Calculation of locus-specific expression ... run-ClassI-class.bowtielog Traceback (most recent call last): File "/mnt/data/software/seq2HLA/seq2HLA.py", line 791, in main() File "/mnt/data/software/seq2HLA/seq2HLA.py", line 786, in main pipe.run() File "/mnt/data/software/seq2HLA/seq2HLA.py", line 140, in run self.call_HLA( File "/mnt/data/software/seq2HLA/seq2HLA.py", line 252, in call_HLA self.expression(locus_list, length_dict, map, run_name) File "/mnt/data/software/seq2HLA/seq2HLA.py", line 694, in expression totalreads = float(linecache.getline(logfile, 1).split(':')[1]) IndexError: list index out of range bash: 002/hla_map/tempus/R_2.fastq: No such file or directory

katahdin0 commented 1 month ago

I installed the version of bowtie used to develop seq2HLA, so it seems to have moved past this error, but now fails at another step, with an equally opaque error, basically this never gets to the class II genotypes, only class I gets some stuff , I reverted to old version of bowtie and seq2HLA still fails after a while with this message , is this about python 2 versus python 3 ? The digital haplotype is written into /mnt/data/pfo/PFO002/hla_map/tempus/seq2HLA-ClassI-nonclass.digitalhaplotype3 Took 0.5703256130218506 seconds Traceback (most recent call last): File "/mnt/data/software/seq2HLA/seq2HLA.py", line 791, in main() File "/mnt/data/software/seq2HLA/seq2HLA.py", line 786, in main pipe.run() File "/mnt/data/software/seq2HLA/seq2HLA.py", line 152, in run self.call_HLA( File "/mnt/data/software/seq2HLA/seq2HLA.py", line 304, in call_HLA if fourDigit_solutions1[locus] > 1: TypeError: '>' not supported between instances of 'str' and 'int'

katahdin0 commented 1 month ago

Installing the old version of bowtie (0.12.7) and using python 2.7 fixed most problems, but one problem appeared in one dataset and that was due to a mistake in line 306 of seq2HLA, the call to get_max_P should be replaced with self.get_max_P and now there have been no more errors

katahdin0 commented 1 month ago

so old version of bowtie (0.12.7), and python 2.7 with the appropriate biopython (pip install biopython==1.76) and a modification in seq2HLA (line 306 replace get_max_P with self.get_maxP) and the problems disappeared, at least for me