Problems With Mirdeep2

Question

[pmuench]

I want to find miRNAs on a fastq file with mirDeep2. After adapter clipping and alignment (with mapper.pl from mirDeep2) against human_g1k_v37 (indexed with bowtie) I get following error message:

Error: Genome file /PathToFile/human_g1k_v37.fasta has not allowed whitespaces in its first identifier

The first line of the reference file looks like this:

>1 dna:chromosome chromosome:GRCh37:1:1:249250621:1

If I delete all white spaces in the identifiers, and rerun the mapping and mirDeep2, I get an other error message:

The mapped reference id 1 from file exmaple_collapsed.arf is not an id of the genome file /PathToFile/human_g1k_v37.fasta

Any idea how i have to convert the reference file? Thank you!

Answer 1

[JC]

You need to rename the sequences in your reference fasta, just remove everything after the space:

 perl -plane 's/\s+.+$//' < genome.fa > new_genome.fa

You will need to create a new index.

[Gabe Anderson]

Hi,

I know this post has been out for a while but I'd appreciate some help. I tried a few data sets on mirdeep2 and I always got this error:

First line of FASTA reads file is not in accordance with the fasta format specifications
Please make sure your file is in accordance with the fasta format specifications and does not contain whitespace in IDs or sequences

***** Please check if the option you used (options c) designates the correct format of the supplied reads file earthwormShort1.fa *****

I have tried to remove whitespaces several times to ensure my fasta file is OK but the error keeps returning. Please, kindly help.