Dear colleagues, what is the best way (I am preferring a user-friendly workflow) to find the homologous sequences in my FASTA dataset? Imagine, I have a dataset of 100 protein sequences from different organisms and want to find if some of them are mutually homologous/similar without any previous knowledge or clues about what I am looking for. BLAST and similar tools are useless here because I have no single query sequence to submit...
Martin,
you don't need a single query for BLAST. You can use the 100 sequences both as query and subject, go for tabular output and then analyze the results by conventional means.
There is a more specialized protocol to the one you require, called "reciprocally best hits analysis" to find ortologues that may be of interest in this context. See e.g. Article Quickly finding orthologs as reciprocal best hits with BLAT,...
I don't know any special software but the task is not that difficult to script or searching for RBH may give you an interesting answer as well.
Dear Matej,
thank you for your prompt answer, this sounds like a great solution, I will try :)
Thanks
Martin
Dear Martin,
A simple and quick way to carry out your analysis would be to do it from inside the terminal application on your computer or on a server on which the BLASTp program already has been installed. By using the Unix or Linux command line you should first format your fasta file containing 100 sequences to a BLASTp database with the following command:
makeblastdb -in your_fasta_file -dbtype prot -parse_seqids -out your_blast_database -title "my 100 proteins"
Once your database has been formatted, you do an "all against all" BLASTp search of each of the 100 sequences against all the other 99 sequences by typing the following command:
blastp -db your_blast_database -num_alignments 0 -num_descriptions 2 -evalue 1e-20 -query your_fasta_file.
The output of this command will be a long list of each of your 100 sequences with itself and with its single best orthologue, an output easy to analyze. Good luck.
Dear Professor Fred R. Opperdoes,
really appreciate your help and detailed answer. Although I am not a Linux user, I believe that your protocol will be applicable also to BLAST+ running on my Windows, I will try :)
All the best,
Martin
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts