HISAT2 alignment problem

Question

[kwanghoon]

I have problems when I run hisat2. my command is

./hisat2 -x $HISAT2G/hisat2_hg38 -p 16 -1 /home/sunghyepark_lab/test/test_files/RNA/rawData/trimmed/trimmed_paired_1607370-RNA-T_1.fastq.gz -2 /home/sunghyepark_lab/test/test_files/RNA/rawData/trimmed/trimmed_paired_1607370-RNA-T_2.fastq.gz -S /home/sunghyepark_lab/test/test_files/RNA/rawData/HISAT2_aligned/1607370_RT.sam

used reference hg38 indexed by hisat2-build.

I got errors like

File "/home/sunghyepark_lab/packages/hisat-genotype-top/hisat2_read_statistics.py", line 182 length_map = sorted(length_map.iteritems(), key=lambda (k,v):(v,k), reverse=True) SyntaxError: invalid syntax

so I googled this error, and changed code "key=lambda (k,v):(v,k) to key=lambda k,v:(v,k)" no error message like above. But another error came out...

File "/home/sunghyepark_lab/packages/hisat-genotype-top/hisat2_read_statistics.py", line 183 print cnt, mn, mx, avg, ",".join([str(k) for (k,v) in length_map]) SyntaxError: invalid syntax

help me please..

Let me know if any other informations are needed to solve this problem

[ATpoint]

Output of python --version?

[kwanghoon]

version is python --version Python 3.7.6

[geneticatt]

I'm getting the same error when running hisat2 (also as a batch submission script). I don't have permissions to edit the script on our cluster. For the case that it does need to be fixed, is there a way to copy the script and provide its location in the batch script so that hisat is provided the alternative (edited) script when hisat2_read_statistics.py is called?

I have a feeling that fixing the script is not necessary because the read mapping seems to be taking place anyways:

File "/isg/shared/apps/hisat2/2.2.0/hisat2_read_statistics.py", line 182
    length_map = sorted(length_map.iteritems(), key=lambda (k,v):(v,k), reverse=True)
                                                           ^
    SyntaxError: invalid syntax
    10070864 reads; of these:
      10070864 (100.00%) were paired; of these:
        831566 (8.26%) aligned concordantly 0 times
        8512855 (84.53%) aligned concordantly exactly 1 time
        726443 (7.21%) aligned concordantly >1 times
        ----
        831566 pairs aligned concordantly 0 times; of these:
          26687 (3.21%) aligned discordantly 1 time
        ----
        804879 pairs aligned 0 times concordantly or discordantly; of these:
          1609758 mates make up the pairs; of these:
            1008340 (62.64%) aligned 0 times
            465483 (28.92%) aligned exactly 1 time
            135935 (8.44%) aligned >1 times
    94.99% overall alignment rate

I imagine it's poor practice to let a module of your program break and continue to run it anyways, so ideas for solutions would be greatly appreciated. I have contacted the admins requesting that they edit the script, but this process usually takes a week or so, which is time I can't spare for this analysis.

[ATpoint]

Just get your own version of the tool using a package manager like miniconda that does not require root permission.

Answer 1

[ATpoint]

Yeah, this seems to be a problem related to python3 whereas the script expects python2.

You can probably solve this by modifying the header of that script hisat2_read_statistics.py to:

#!/usr/bin/env python2

as this makes it explicit that python2 is to be used.

[kwanghoon]

Use of uninitialized value $read_len_str in string ne at ./hisat2 line 293.

86563369 reads; of these:

86563369 (100.00%) were paired; of these:

1516202 (1.75%) aligned concordantly 0 times
55773439 (64.43%) aligned concordantly exactly 1 time
29273728 (33.82%) aligned concordantly >1 times
----
1516202 pairs aligned concordantly 0 times; of these:
  135333 (8.93%) aligned discordantly 1 time
----
1380869 pairs aligned 0 times concordantly or discordantly; of these:
  2761738 mates make up the pairs; of these:
    1048277 (37.96%) aligned 0 times
    1011554 (36.63%) aligned exactly 1 time
    701907 (25.42%) aligned >1 times
  

99.39% overall alignment rate

I think it worked! Thank you

But I don't know what Use of uninitialized value $read_len_str in string ne at ./hisat2 line 293. means

Is this not the big problem?