I have a dataset of 5640 protein sequences in fasta format. I want to select and make a new dataset with the proteins having more than 100 amino acids. How can I do it using cmd command or Rstudio? What should be the script to do this? I badly need the answer!! Please help me.....
For fitlering sequences with minimum 100 aa/nt length i.e > 100,
$ seqkit seq -m 100 test.fa
For filtering sequences with maximum of 1000 aa/nt length<=1000
$ seqkit seq -M 1000 test.fa
For filtering sequences between maxiumum size 200 and minimum size of 100:
$ seqkit seq -m 100 -M 200 test.fa
Download seqkit from here
what is OS version you are on and where is seqkit located (folder)? If you are on windows 10, you can use linux along with windows 10. (https://docs.microsoft.com/en-us/windows/wsl/install-win10). It is a long procedure. However it makes life easier. Once you install linux, try installing conda in linux. Conda/bioconda has most of the bioinformatics software and easy to install them.
For windows 7 and 10:
shift button and right click on the same time in open area (white space or on freespace on desktop). Not on any file or directory.seqkit.exe version. It should show seqkit version some thing like seqkit v0.8.0..dir in the terminal. It should show all the files. To list .fa files, type dir *.fa. It should show only .fa only. This should list your fasta file here.seqkit.exe seq -m 100 test.fa -o output.fa (replace test.fa with your file namem output.fa with output file name you want). For ubuntu:
seqkit version in the terminal.cd ~/Desktop".ls *.fa or ls.fasta or ls.fna depending on the the extension of fasta file. Seqkit can handle gzipped fasta file as well.seqkit seq -m 100 test.fa -o output.fa (replace test.fa with your file namem output.fa with output file name you want). 
Here's an awk-based approach, if your FASTA records are single-lined (header on one line, sequence on the next line, and so on):
$ awk '!/^>/ { next } { getline seq } length(seq) > 100 { print $0 "\n" seq }' input.fa > output.fa
If multi-lined (header on one line, sequence split across two or more lines), then the input.fa would need preprocessing to make it single-lined. There are other answers on Biostars that explain how to do this.
Via: http://itrylinux.com/use-awk-to-filter-fasta-file-by-minimum-sequence-length/
Explanation in that page is incorrect . Code can be shortened further, some thing like this:
$ awk '/^>/ { getline se } length(se) >100 { print $0 "\n" se }' test.fa
Example fasta: (copy/pasted from the page in the link):
$ cat test.fa
>sequence_ID_1
atcgatcgggatcaatgacttcattggagaccgaga
>sequence_ID_2
gatccatggacgtttaacgcgatgacatactaggatcagat
Ouptut from the code in linked page:
$ awk '!/^>/ { next } { getline seq } length(seq) <37 { print $0 "\n" seq }' test.fa
>sequence_ID_1
atcgatcgggatcaatgacttcattggagaccgaga
output from the in this post:
$ awk '/^>/ { getline se } length(se) < 37 { print $0 "\n" se }' test.fa
>sequence_ID_1
atcgatcgggatcaatgacttcattggagaccgaga
Length of the sequences:
$ awk '!/^>/ { print length($NF) }' test.fa
36
41
Is it incorrect?
$ awk '!/^>/ { next } { getline seq } length(seq) > 35 { print $0 "\n" seq "\n" length(seq) }' /tmp/in.fa
>sequence_ID_1
atcgatcgggatcaatgacttcattggagaccgaga
36
>sequence_ID_2
gatccatggacgtttaacgcgatgacatactaggatcagat
41
Seems like the filter works:
$ awk '!/^>/ { next } { getline seq } length(seq) > 36 { print $0 "\n" seq "\n" length(seq) }' /tmp/in.fa
>sequence_ID_2
gatccatggacgtttaacgcgatgacatactaggatcagat
41
Am I missing something?
Code is correct and uses double negation. Explanation in the page is incorrect.
!/^>/ { next } { getline seq }
!/^>/ - get lines that do not start with > (i.e sequences in fasta).{ next } - get next lines of lines that do not start with > (i.e sequence headers)getline - get next lines next to the lines that do not start with > (i.e sequences in fasta- back to step1)Instead it may have been :
/^>/ { getline seq }
/^>/ - get lines with > (i.e headers)get line - get lines next to lines start with > (i.e sequences)Explanation in the page for !/^>/ {next} is that: (copy/pasted)
If a line (i.e. record) begins with a “>”, go to the next line (record).
You can use SEDA (http://www.sing-group.org/seda/), which has an option to filter sequences by length (http://www.sing-group.org/seda/manual/operations.html#filtering) as well as other types of filtering.
Regards.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
What format is your data in already? Have you created a dataframe?
Please provide more information - ideally a minimal representative dataset (the first 10 lines of a dataframe with
head()would do for instance).I haven't created a separate dataframe. The data are in fasta format in a fasta file. For example, below are the first 10 sequences of this dataset........
I just want to use a command to select the protein sequences having 100 or less than hundred amino acid and exclude them from the dataset and get a new dataset containing remaining proteins having >100 amino acid.
For an R solution, see https://stackoverflow.com/questions/8640377/remove-all-rows-where-length-of-string-is-more-than-n for a hint.
Your data is in a horrible, horrible format at the moment however. You've no easy way of demarcating the headers from the sequences as they're separated by a single whitespace and both your headers and sequences also have spaces in.