can't work with DNA sequences

[jun3234]

I can run scripts with default protein sequences. But, when I ran scripts with dna sequences, likes:

python2 repset.py --outdir test --seqs test.fa

And, after database was maked, it occurs error.

2020-08-17 18:46:36,841 INFO:makeblastdb -in test.fa -input_type fasta -out test/db -dbtype prot

Building a new DB, current time: 08/17/2020 18:46:36
New DB name:   /home/smrtanalysis/tools2/submodular_sequence_repset/test/db
New DB title:  test.fa
Sequence type: Protein
Keep MBits: T
Maximum file size: 1000000000B
Adding sequences from FASTA; added 88170 sequences in 3.99215 seconds.

2020-08-17 18:46:40,911 INFO:psiblast -query test.fa -db test/db -num_iterations 6 -outfmt 6 qseqid sseqid pident length mismatch evalue bitscore -seg yes -out test/psiblast_result.tab
BLAST engine error: Could not calculate ungapped Karlin-Altschul parameters due to an invalid query sequence or its translation. Please verify the query sequence(s) and/or filtering options
Traceback (most recent call last):
  File "repset.py", line 1585, in <module>
    db = run_psiblast(workdir, args.seqs)
  File "repset.py", line 98, in run_psiblast
    subprocess.check_call(cmd)
  File "/home/smrtanalysis/python2/lib/python2.7/subprocess.py", line 186, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['psiblast', '-query', path('test.fa'), '-db', path('test/db'), '-num_iterations', '6', '-outfmt', '6 qseqid sseqid pident length mismatch evalue bitscore', '-seg', 'yes', '-out', path('test/psiblast_result.tab')]' returned non-zero exit status 3

When I mod the scripts, chane prot to nucl in line 19, But, it also complains.

2020-08-17 19:00:09,751 INFO:makeblastdb -in test.fa -input_type fasta -out test/db -dbtype nucl

Building a new DB, current time: 08/17/2020 19:00:09
New DB name:   /home/smrtanalysis/tools2/submodular_sequence_repset/test/db
New DB title:  test.fa
Sequence type: Nucleotide
Keep MBits: T
Maximum file size: 1000000000B
Adding sequences from FASTA; added 88170 sequences in 3.5302 seconds.

2020-08-17 19:00:13,328 INFO:psiblast -query test.fa -db test/db -num_iterations 6 -outfmt 6 qseqid sseqid pident length mismatch evalue bitscore -seg yes -out test/psiblast_result.tab
BLAST Database error: No alias or index file found for protein database [test/db] in search path [/home/smrtanalysis/tools2/submodular_sequence_repset::]
Traceback (most recent call last):
  File "repset.py", line 1585, in <module>
    db = run_psiblast(workdir, args.seqs)
  File "repset.py", line 98, in run_psiblast
    subprocess.check_call(cmd)
  File "/home/smrtanalysis/python2/lib/python2.7/subprocess.py", line 186, in check_call
    raise CalledProcessError(retcode, cmd)
subprocess.CalledProcessError: Command '['psiblast', '-query', path('test.fa'), '-db', path('test/db'), '-num_iterations', '6', '-outfmt', '6 qseqid sseqid pident length mismatch evalue bitscore', '-seg', 'yes', '-out', path('test/psiblast_result.tab')]' returned non-zero exit status 2

And, how can I run this script with dna sequences?

Tanks~ Jun

[mlibbrecht]

That's strange, I haven't encountered that problem before. Does it work if you run the makeblastdb and psiblast commands on the command line yourself? Otherwise, you might try looking on the psiblast troubleshooting pages.

[jun3234]

There those files are in current dircetory.

[smrtanalysis@localhost submodular_sequence_repset]$ ll -htr
total 154M
-rw-r--r--. 1 smrtanalysis xialab 1.0K Aug  7 10:36 Readme.md
-rw-r--r--. 1 smrtanalysis xialab  78M Aug 17 18:31 final.unaligned.fa
drwxr-xr-x. 2 smrtanalysis xialab  189 Aug 17 18:31 out
-rw-r--r--. 1 smrtanalysis xialab  76M Aug 17 18:40 test.fa
-rw-r--r--. 1 smrtanalysis xialab  63K Aug 17 18:59 repset.py
drwxr-xr-x. 2 smrtanalysis xialab  189 Aug 17 19:00 test

And, I checked commands in stdout.txt.

[smrtanalysis@localhost submodular_sequence_repset]$ cat ./test/stdout.txt
2020-08-17 19:00:09,751 INFO - makeblastdb -in test.fa -input_type fasta -out test/db -dbtype nucl
2020-08-17 19:00:13,328 INFO - psiblast -query test.fa -db test/db -num_iterations 6 -outfmt 6 qseqid sseqid pident length mismatch evalue bitscore -seg yes -out test/psiblast_result.tab

Then, I ran this commands in stdout.txt manually.

[smrtanalysis@localhost submodular_sequence_repset]$ makeblastdb -in test.fa -input_type fasta -out test_man/db -dbtype nucl

Building a new DB, current time: 08/17/2020 23:10:50
New DB name:   /home/smrtanalysis/tools2/submodular_sequence_repset/test_man/db
New DB title:  test.fa
Sequence type: Nucleotide
Keep MBits: T
Maximum file size: 1000000000B
Adding sequences from FASTA; added 88170 sequences in 4.24149 seconds.

[smrtanalysis@localhost submodular_sequence_repset]$ psiblast -query test.fa -db test_man/db -num_iterations 6 -outfmt '6 qseqid sseqid pident length mismatch evalue bitscore' -seg yes -out test_man/psiblast_result.tab
BLAST Database error: No alias or index file found for protein database [test_man/db] in search path [/home/smrtanalysis/tools2/submodular_sequence_repset::]

But, Error occurs again. Do I need build protein database? While I just have dna sequences.

[mlibbrecht]

Everything in repset should work with nucleotide sequences. I'm not sure why psiblast isn't working, and I don't have any experience with that bug, sorry -- you should try asking on the psiblast troubleshooting.