NCBI ORFfinder - How Does It Work

Question

[A Soggy Waffle]

NCBI help desk hasn't been very helpful to me on this so far.

I'm looking for any record or docs on the ORFfinder tool. I want to know how it finds ORFs. Not how to use it, but how it determines what is and is not an ORF. Is it regex, HMM, fairy dust? Maybe such docs are posted somewhere but I can't seem to find them.

Thanks

[GenoMax]

Download the program from here. gunzip it and add execute permissions chmod a+x ORFfinder. Run it to look at in-line help.

[A Soggy Waffle]

I guess I wasn't clear enough. It's not the arguments and calling the program that's the problem. I want to know how it finds ORFs, algorithmically.

[Joe]

If this is the ORF finder I think it is, I don’t think it’s much more sophisticated than finding start and stop codons in all 3 frames on each strand, and highlighting the interval as a potential ORF. Possibly some filtering for minimal sizes.

‘Proper’ gene predictors like Prodigal and Glimmer have much more sophisticated characteristics like sequence complexity, homology, etc.

[GenoMax]

See my note below.

[GenoMax]

Video above contains an email address of a person who probably still supports this tool. You can try writing to them directly. It does not contain the information you are looking for.

Note 1: ORFfinder is not meant to be a gene prediction tool for eukaryotes.
Note 2: NCBI help desk can take up to 3 working days to respond. Likely because of volume of queries they get. I have always received an answer for my questions.

Answer 1

[h.mon]

ORF Finder perform six-frame translation of a nucleotide sequence given a particular genetic code, finding all ORFs possible. It then filters for ORFs above a certain size threshold, and may exclude ORFs nested inside longer ORFs. This information (and nothing more) can be found at the paper Database resources of the National Center for Biotechnology.