ORF finder script

Question

[ahm3dhany]

I wrote a basic bash script to find the ORF (i.e. open reading frame) in a given nucleotide sequence and I need to know if I done it right.

ORF_finder.sh :

#!/bin/bash

var=$1
grep -Eo --color=auto 'ATG(...)*T(A(A|G)|GA)' $1

an example for the usage:

~$ sequence="CTGGATGATCCTCTAACCGCGCAAACGAGG"
~$ echo $sequence | ./ORF_finder.sh
ATGATCCTCTAA

another example:

~$ sequence="ATAATCGGCCTCGACATCCTCCGCCACGAAAGGACTGTCCCCAATCCCGAAGGCCGCAGAGCGCTGCACATACTAGGGTCAGCTAAGGTACCTCTCATGCGAGGACACGCGATGTGGCATCTAGGCGGCGTTAGAAAATTATTCGAGGCGGCCTACCGTCTTAACCGTTAAATACAAGCATGGGAAGGCAGAGCGAAAATAAAATTGCCCGCGCCTCACTACCTGCCGTCTCGTAACACTTAGCTCTAAAATAGAGTAAGCTCGGCCCCCAGTCCAAGGCACGTAAGGATGTATCGAGGCTCAAAAGACTCGCTGATCGTACCGGTCTCGTGCGTAAAAAGGCAGCAGAACTATGCTTGACTATCCATACGTCTCCATCGTTCCTGCTGATTCGTCGCGAATTGGCGCGGTTACTTAGTCTCCGGGCTGTCCGGTCGGGCTAGGTGATGCCTGTCCCTAAGGTGAATCAAGAAATCCTCAAAACTGCATAATCACGTGTT"
~$ echo $sequence | ./ORF_finder.sh
ATGCGAGGACACGCGATGTGGCATCTAGGCGGCGTTAGAAAATTATTCGAGGCGGCCTACCGTCTTAACCGTTAAATACAAGCATGGGAAGGCAGAGCGAAAATAAAATTGCCCGCGCCTCACTACCTGCCGTCTCGTAACACTTAGCTCTAAAATAGAGTAAGCTCGGCCCCCAGTCCAAGGCACGTAAGGATGTATCGAGGCTCAAAAGACTCGCTGATCGTACCGGTCTCGTGCGTAAAAAGGCAGCAGAACTATGCTTGACTATCCATACGTCTCCATCGTTCCTGCTGATTCGTCGCGAATTGGCGCGGTTACTTAGTCTCCGGGCTGTCCGGTCGGGCTAGGTGATGCCTGTCCCTAAGGTGA

Is finding ORF that simple? or it's more complicated than that.

[cschu181]

If you only want ORFs from the forward frames and which are not supported by any evidence, then, yes, it is that simple.

[Joe]

Yup, if your definition of ORF is sufficient that all you care about is it starts with an ATG (they don't always) and ends with one of the stop codons some multiple of 3 away, then yeah, it's as simple as that.

More sophisticated ORF finders will consider the 6 possible reading frames (forward and reverse), as well as possibly include a minimum length, and some filtering for sequence complexity etc.

It looks like this will only work if you chomp the newlines in a fasta too beforehand - I think...

[ahm3dhany]

thanks.. could you please provide me with an article or paper or anything that elaborate the other details I neglected.

[Joe]

One of the most sophisticated tools for ORF detection is GLIMMER: https://ccb.jhu.edu/papers/glimmer2.pdf

That is even more complex than just the things I mentioned though. It implements Markov models. NCBI's ORF finder is slightly more complex than basic string searching, however I don't know exactly what the code is doing. You might be able to find it on the web somewhere to download:https://www.ncbi.nlm.nih.gov/orffinder

[WouterDeCoster]

In addition, only one ORF per fragment...

[ahm3dhany]

First of all, thank you for your answer.. do you mean by "forward frames" that if each nucleotide of the sequence flipped (i.e. A->T, T->A, C->G and G->C) ?

[WouterDeCoster]

The reverse complement of the sequence.