My (Max's?) Minimal Fasta Toolkit

Nothing sophisticated. Minimal, simple fasta tools. Each program is self-contained in the ./src/fasta directory, and follows similar boilerplate code, related to file handling.

Usage

Typing mmft (shows subcommands) or mmft <subcommand> -h (shows specific subcommand) will show the usage of the tool in question.

Commands are added only as and when I need them. If you like what you see, please feel free to contribute a PR with your favourite subcommand.

Calculations

• mmft len <fasta(s)> or cat <fasta(s)> | mmft len. Calculates lengths of each fasta record.
• mmft gc <fasta(s)> or cat <fasta(s)> | mmft gc. Calculates GC content of each fasta record.
• mmft n50 <fasta(s)> or cat <fasta(s)> | mmft n50. Calculates n50 of a fasta record (or stream of fasta files combined).
• mmft num <fasta(s)> or cat <fasta(s)> | mmft num. Calculates number of sequences, and total number of base pairs in the fasta file input(s).

File manipulations

• mmft regex -r "<regex>" <fasta(s)> or cat <fasta> | mmft regex -r "<regex>". Extracts fasta records from one or multiple fasta files with headers matching the regex.
• mmft extract -r 1-100 <fasta(s)> or cat <fasta> | mmft extract -r 1-100. Extracts first 100 nucleotides from each fasta record. You can of course choose any range, using a dash to separate the numbers.
• mmft filter -f <file> <fasta(s)>. Supply a text file of one ID per line and filter will extract the corresponding fasta records.
• mmft merge <fastas>. Will merge multiple fasta files together into the same record.
• mmft sample <fasta(s)> -n <N>. Will randomly sample a fasta file (or stream of fasta files) to a specified number of records.

Careful when piping into mmft as fasta files are not treated separately, they are treated as a continuum of fasta records. Hence, while mmft n50 1.fasta 2.fasta shows the n50 of each fasta file separately, cat *.fasta | mmft n50 will calculate the n50 of both files combined. In addition, mmft sample loads the entire STDIN into memory, so be careful when piping large files.

All printed to STDOUT.

TODO's

I'll add stuff as and when I have time, or they are of use. Maybe:

• Simple pattern matching, returning positions.
• Potential ORFs
• Any kmer stuff?
• Testing
• Better documentation