Calculating dinucleotide frequencies for each sequence in a multi fasta file

Question

[emilyc]

Bonjour,

I am looking for an easy way to calculate the dinucleotide frequencies of every sequence in a multi fasta file, and to have these saved with the sequence names associated with each frequency. Is there an easy command, tool, or way to do this using bash, or R, etc.?

I can think of some complicated methods of doing this, but if something is already programmed to do this then that would be amazing.

Merci pour votre temps,

Emily

[Andrzej Zielezinski]

Jellyfish is a popular tool for counting kmers in a multi-fasta file.

[GenoMax]

Script To Calculate Dinucleotide Frequency For Many Sequences
Biostrings package in R (see page 72).

[emilyc]

My understanding of this was that it's calculating the dinucleotide frequencies for all of the sequences over all, and not each. Am I incorrect in thinking this? I did not test this one but the OP said "I should note that I do not need to calculate each sequence's dinucleotide frequency. I need to get the di-nt frequency from all sequences. Thank you for your attention." in a comment.

Thank you.

[GenoMax]

You can iterate through your sequences as input and get those for each sequence. Or are you asking something else?

[emilyc]

I mean yeah I can do this in a round about way, however, I am posting here to see if someone has/knows of a nicely done tool that fits my needs already with some organised outputs such as a dataframe with all frequencies for each seq with the original ID, etc. I can do the work, but if I don't need to then that is helpful.

Answer 1

[trausch]

The kent utils from the UCSC genome browser have a rapid tool for this

faCount -dinuc genome.fa

Precompiled binaries are here.

Answer 2

[Joe]

There is a solution in this thread that should work: A long run time problem