Hello I have been trying to run Quake Error Corrector on my Illumina reads before assembly. I am following the manual at http://www.cbcb.umd.edu/software/quake/manual.html. I have a problem running the final step of error correction as Quake always gives an error
"Kmers are not of expected length 21"
As I understand it the Quake program has 3 sub-programs inside it called - 1)Count qmers 2)Coverage Cutoff 3)Correct reads I have tried searching for examples to run Quake but in all of the cases, nobody uses the second program which is Cov_model.py. The Quake.py program is directly run like
quake.py -f pairs.list -k 21
However, there are no commandline options for cut-off value or for providing a count-file.
My commands in brief were-
1) Using jellyfish to get kmer counts.
jellyfish count -m 21 -s 27G -t 48 -C -F 2 File_1.fastq File_2.fastq
This generated a file called mer_counts.jf
2)The second module in jellyfish computes a histogram of k-mer ocurrences.
jellyfish histo -f -t 48 -v -o reads.histo mer_counts.jf
This generated a file called reads.histo containing number of kmers and their multiplicity
3)Then I can plug it into Quake and use
cov_model.py --int reads.histo
This worked well and gave a Coverage Cutoff of 1.
4) I went ahead using reads.histo to the final Correction step
correct -f input_files.txt -k 21 -c 1 -m reads.histo -p 48 -z --log
This just gives me an error saying "Kmers are not of expected length 21"
Should I just go ahead with using Quake.py? It does not provide any options to provide a cutoff file or a count file. Please advise!
