how do KOG classification and KEGG analysis based on nucleotide sequences?

Question

[Kurban]

Hello guys,

I have more than 400 up regulated transcripts sequences from my insect transcriptome data. If I wanna do KOG classification and KEGG analysis according to their sequences how should I do that? Could u give me some tips?

Thanks in advance

[5heikki]

Have you read this?

[Kurban]

thanks @5heikki,

I downloaded Kog_LE.tar.gz file from ftp://ftp.ncbi.nih.gov/pub/mmdb/cdd/little_endian/ and did rpsblast:

rpsblast -i golf-4-down-and-DEseq-down-intersection.fasta -d Kog -m 2 -e 0.01 -a 3 -o golf-4-down-and-DEseq-down-intersection_Kog_results

and got the result:

in query file, there are 60 sequences, but most of them aligned to gnl|CDD|231029 (Nuclear pore complex) with lowest e-value and highest score.my questions here are: did I do KOG classification correctly? if I did what is the proper interpretation of the result file? or is there something else I should do for making the rpsblast results little more understandable?

[Kurban]

thanks @Michael Dondrup,

yeah, i did used the nucleotide sequences as query file and after i added the -p argument that worked fine.

hey Micheal, when you do Kog what e-value do u use to get an acceptable results?

[Michael]

1e-6 is often used, but these alignments can be short and bad, 1e-10 is a more conservative cutoff. There are many sequenced insects and Drosophila is a model organism. I'd expect to see a lot of hits with 1e-50 or better.

Answer 1

[Michael]

It seems like you did an untranslated nucleotide search with rpsblast against a protein db, so your result is bogus. You are using legacy rpsblast which had an option:

 -p  Query sequence is protein  [T/F]  Optional
    default = T

Try to set this to F, then it might do the translation.

For Kegg analysis you should use KAAS which can handle nucleotide input.