Hi, I am facing again issues with Diamond in versi...
# taxprofiler
s
Hi, I am facing again issues with Diamond in version 1.2.2, Although diamond ran successfully the number of reads assigned are very low (21). But with the version 1.8.1 I got 914592. Any clues?
j
IS this with paired-end data?
s
yes
j
And did you merge your reads?
When you say version 1.8.1 you mean you ran DIAMOND manually?
s
yes using option --shortread_qc_mergepairs, no with Taxprofiler 1.1.8 version and Taxprofiler 1.2.2 version
j
OK the only thing we have changed in regards to DIAMOND is the versio nof DIAMOND: 2.0.15 to 2.1.8
One thing that @Lili Andersson-Li pointed out is to check how many reads you have after merging
And did you use the
--shortread_qc_includeunmerged
flag?
OK the only thing we have changed in regards to DIAMOND is the versio nof DIAMOND: 2.0.15 to 2.1.8
So you could try and see if there were any major changes to the tool itself
s
I see the reads after nmerging reduced to Kilobases from megabases, no I didnot use --shortread_qc_includeunmerged
j
If your reads are too long for merging and don't include the unmerged reads that maybe why you had so few reads teh first time around
I'm not sure why you would get so many more in the latest version unless you changed yoru database?
s
I get less reads in the latest TP version, may be due to missing lot of reads on merging I will try using thos flag --shortread_qc_includeunmerged
j
Be careful though:
If you have a mixture of merged and unmerged reads you may get imbalanced taxon counts
(@Lili Andersson-Li is working on a fix for this now)
s
Ah ok thats nice, could you please let me know which tool is used for merging I could try to run the command locally and see to it
j
You can check the parameters page on the docs or in your MultiQC
IT depends what option you selected
s
ok I will check that, Thank you