Dear Sunmathi,

For multiple sequences, I just can recommend to use RNAalifold, ont the web server: http://rna.tbi.univie.ac.at/cgi-bin/RNAWebSuite/RNAalifold.cgi for example.

Okay Sir. I will try for it. 

RNAalifold will actually get a consensus sequence and predict its secondary structure I guess. But Sir, I need to predict structure for all the sequences automatically without manually uploading each single sequence to RNAfold. Is there any way to do it sir?

Hi,

Insa Kruse's answer is the way to go. I just had to do this. Here are some tips for it to work since I could not import the RNAfold module into python.

After you install RNAfold from ViennaRNA, open python3 and see if you can import the module RNA (import RNA). If it fails, which it did for me, go to the following location (you may need to change the python3 to the version you have): "/usr/local/lib/python3.6/site-packages/RNA" and rename the .so file to _RNA.so so that python can recognize it and I also had to add it to the system path (see below).

Here is the code I used to get the folding energy (I have biopython installed):

import sys

sys.path.append("/usr/local/lib/python3.6/site-packages/RNA")

import _RNA as RNA

from Bio import SeqIO

import csv

## read fasta file called seqs.fa

records = list(SeqIO.parse("seqs.fa", "fasta"))

## make empty lists for collecting the name

## of the fasta record and the folding energy

names = []

mfes = []

## iterate on the records and get information

for i in range(0,len(records)):

nm=records[i].name

seq=records[i].seq

seq=str(seq)

(ss, mfe) = RNA.fold(seq)

names.append(nm)

mfes.append(mfe)

## write to csv the results

rows= zip(names,mfes)

with open("RNAfoldMFE.csv", "w") as f:

writer = csv.writer(f)

for row in rows:

writer.writerow(row)