miRDeep2 arf format error

[anibhax]

Hi,

I'm using miRDeep2 to find new miRs in Mouse genome. I have RNA seq data from Illumina.

I used mapper.pl to generate .fa and .arf files. Now I'm using miRDeep2.pl to find new miR's. But it shows that the arf format is incorrect.

Here are the commands I have used:

mapper.pl SampleC.fastq -e -h -p ~/refs/mm9 -s SC.fa -t SC.arf

miRDeep2.pl Sample.fa ~/refs/mm9.fa Sample.arf miRBase_mmu_v20.fa none none -t Mouse

I get this error when I ran miRDeep2.pl:

#####################################
# #
# miRDeep2.0.0.5 #
# #
# last change: 25/06/2012 #
# #
#####################################

miRDeep2 started at 13:57:10


#Starting miRDeep2
#Starting miRDeep2
/opt/sam/miRDeep/2.0.5/miRDeep2.pl SampleD.fa
/home/abc/refs/mm9.fa SampleD.arf ../miRBase_mmu_v20.fa none none -t Mouse

miRDeep2 started at 13:57:10


mkdir mirdeep_runs/run_07_05_2014_t_13_57_10

Error: Mapping file SampleD.arf is not in arf format

Each line of the mapping file must consist of the following fields
readID_wo_whitespaces length start end read_sequence genomicID_wo_whitspaces length start end genomic_sequence strand #mismatches editstring
The editstring is optional and must not be contained
The readID must end with _xNumber and is not allowed to contain whitespaces.
The genomeID is not allowed to contain whitespaces.

This is the arf file of the sample:

[abc mirdeep2]$ head Sample.arf
seq_2 51 1 51 ggcaagggaagaagacttcaggccactgggaaatgcttggttctgttcagt chr4 51 133526049 133526099 ggcaagggaagaagacttcaggccactgggaaatgcttggttctgttcagt + 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_5 51 1 51 gccattcttgcgaaaatcacggcccaaggtctggatatagttattggagac chr2 51 84460474 84460524 gccattcttgcgaaaatcacggcccaaggtctggatatagttattggagac + 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_6 51 1 51 ccagcttttaatacacatatgtacatacatatgtgcataacaaattatagc chr12 51 45602556 45602606 ccagcttttaatacacatatgtacatacatatgtgcataacaaattatagc - 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_7 51 1 51 ttgagcttgaacgctttctttattggtggctgcttttaggcctacaatggt chr10 51 95540393 95540443 ttgagcttgaacgctttctttgttggtggctgcttttaggcctacaatggt + 1 mmmmmmmmmmmmmmmmmmmmmMmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_8 51 1 51 ggtgtccctgggatactcatagcctcagactggttacagagttggggcttt chr15 51 102043274 102043324 ggtgtccctgggatactcatagcctcagactggttacagagttggggcttt - 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_9 51 1 51 accattgtcgtccagagctccgtccacaccatagctccatcccttgccaca chr5 51 53673136 53673186 accattgtcgtccagagctccgtccacaccatagctccatcccttgccaca - 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_11 51 1 51 ctggagtcttggaagcttgactaccctacgttctcctacaatggaccttga chr9 51 78023154 78023204 ctggagtcttggaagcttgactaccctacgttctcctacaatggaccttga + 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_12 51 1 51 tgtggggattgccctgatgctgacaaagtcagggcaggagtcagaagaaat chr15 51 8585628 8585678 tgtggggattgccctgatgctgacaaagtcagggcaggagtcagaagaaat + 0mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_13 51 1 51 ctttctttacaatgaccaagttgagaacactgagattagcgtccacaatgc chrX 51 66859072 66859122 ctttctttacaatgaccaagttgagaacactgagattagcgtccacaatgc + 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm
seq_13 51 1 51 ctttctttacaatgaccaagttgagaacactgagattagcgtccacaatgc chr2 51 75029531 75029581 ctttctttacaatgaccaagttgagaacactgagattagcgtccacaatgc + 0 mmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmmm

I'm not sure what the error is. The arf file looks alright to me.
Does anyone know what the issue is? Any help appreciated. Thanks!

[Hell-Panther]

Had the same problem

I fixed it.

The problem is that in arf file first field (readID) is seq_2 not seq_x2. So you have to change all readID's adding x before number.

And when you use mapper.pl try to add -m option to your commandline, it makes correct files.

Something like this for collapsed.fa :
>seq_7064753_x43185
GGCTGGTCCGATGGTAGTGGGTTATCAGAACT

And vs_genome.arf :

seq_0_x3041305 22 1 22 tcctgtactgagctgccccgag chr8 22 41518002 41518023 tcctgtactgagctgccccgag - 0 mmmmmmmmmmmmmmmmmmmmmm



Good luck!

[anibhax]

Hi.. thanks for the -m tip. I included it and ran the mapper.pl script again. Now, when I run the mirDeep2.pl I get a new error. This time not about the .arf file, but the reference file.

Error: miRNA reference this species file ../miRBase/miRBase_mmu_v20.fa has not allowed whitespaces in its first identifier

Do you know what it means?

[Hell-Panther]

Probably read names contain whitespaces. Something like: >hsa-let-7a-1 MI0000060
Simply replace the whitespace with underscore, so the read name lookse like this >hsa-let-7a-1_MI0000060
If you use linux server, run in commandline

sed -i 's/ /_/g' filename.fa

this replaces whitespace with underscore.