I have ~800 multifasta files (one gene/file) for which I would like to create independent alignments (preferably with MAFFT) automatically.
So instead of typing mafft --retree 1 input_file > output_file ' for each file I am looking for a way where the next fasta file is automatically grabbed and aligned after the previous alignment has finished.
•
written
by
Caroline S •
Use a UNIX for loop maybe?
FILE_1=`ls *.fasta | head -n 1`
FILE_2='ls *.fasta | head -n 2 | tail -n +2`
OUT_PUT=init_output.fasta
mafft -retree 1 "`cat ${FILE_1} ${FILE_2}`" >${OUT_PUT}
for fasta_file in $(ls *.fasta | tail -n +3)
do
IN_PUT=`cat ${OUT_PUT} ${fasta_file}`
mafft --retree 1 ${IN_PUT} > ${OUT_PUT}
done
This should hold the final output in ${OUT_PUT} once done. I ran this in my head though, so take a backup to avoid loss of data owing to my errors in contingency planning.
Ram, I think OP wants to align each of the 800+ FASTA files individually and write the corresponding output. So, there will be 800+ output files. Assuming OP has the FASTA files with .fasta extension, the following script will run MAFFT and write output with suffix .out added to the input file name.
for fasta_file in $(ls *.fasta)
do
mafft --retree 1 $fasta_file > $fasta_file.out
done
I can see how "one gene/file" might lead you to think that there is only one sequence per file. However the OP mentions "multifasta file" which is a file containing multiple sequences in FASTA format. So, the way I understand is OP has 800+ files with each file having multiple related sequences of similar function ("one gene"), probably from different species in FASTA format and they want to align these related sequences within a file. I think OP is a biologist like me :)
Thank you both for your input! The option suggested by Siva is working out! And indeed, I have multiple sequences for one gene in each file; sorry if that wasn't clear.
•
written
by
Caroline S •
Log in to answer this question.
