SpliceMap error: cannot find chromosomes in bowtie index

[jgw]

I'm trying to get SpliceMap up and running, and run into two errors/bugs, one minor and one major.

1) The minor error is that SpliceMap can’t seem to find the Bowtie index files if their directory in run.cfg uses the $HOME variable (eg, “bowtie_base_dir = $HOME/data/genome”); I’ve gotten around that by using a local directory, but it’s odd because other variables (such as where to put the chromosome fasta files or the reads list) work just fine with $HOME in the filepath. Anyone know what the issue is?


2) The major error is that SpliceMap won’t recognize my preconstructed Bowtie index files, even when I’m using the ones SpliceMap itself generated on a previous run. It keeps giving me the following error:

temp directory exists
Scaning genome: ./zea_mays.chr*.fasta
List of chromosomes to be searched:
chromosome:AGPv2:10:1:150189435:1 | ./zea_mays.chr10.fasta | pos:49 - 152692642
chromosome:AGPv2:1:1:301354135:1 | ./zea_mays.chr1.fasta | pos:47 - 306376751
chromosome:AGPv2:2:1:237068873:1 | ./zea_mays.chr2.fasta | pos:47 - 241020068
chromosome:AGPv2:3:1:232140174:1 | ./zea_mays.chr3.fasta | pos:47 - 236009224
chromosome:AGPv2:4:1:241473504:1 | ./zea_mays.chr4.fasta | pos:47 - 245498110
chromosome:AGPv2:5:1:217872852:1 | ./zea_mays.chr5.fasta | pos:47 - 221504114
chromosome:AGPv2:6:1:169174353:1 | ./zea_mays.chr6.fasta | pos:47 - 171993973
chromosome:AGPv2:7:1:176764762:1 | ./zea_mays.chr7.fasta | pos:47 - 179710889
chromosome:AGPv2:8:1:175793759:1 | ./zea_mays.chr8.fasta | pos:47 - 178723702
chromosome:AGPv2:9:1:156750706:1 | ./zea_mays.chr9.fasta | pos:47 - 159363265
chromosome:AGPv2:UNKNOWN:1:7140151:1 | ./zea_mays.chrUNKNOWN.fasta | pos:57 - 7259211
chromosome:AGPv2:chloroplast:1:140384:1 | ./zea_mays.chrPt.fasta | pos:64 - 142788
chromosome:AGPv2:mitochondrion:1:569630:1 | ./zea_mays.chrMt.fasta | pos:68 - 579192
ERROR: The following chromosome does not exist in the bowtie index
chromosome:AGPv2:10:1:150189435:1 | ./zea_mays.chr10.fasta | pos:49 - 152692642
Please make sure that all chromosomes are contained within the bowtie index

I removed chromosome 10's fasta file, and it gave me the same error for chromosome 1, so it seems to have trouble with all of them.

The relevant parts of run.cfg are

genome_dir = ./
mapper = bowtie
chromosome_wildcard = zea_mays.chr*.fasta
bowtie_base_dir = genome

The genome in question was built by SpliceMapper itself on a previous run, using the exact same fasta files. I don't want to have to rebuild the index every time I run the program; any ideas what's going wrong?

Thanks.

Edit: I also used bowtie-inspect to check the genome files in question, and they do contain the names of all the correct chromosomes.

[paulorapazote]

Hi

The Bowtie index problem could be related with the header of reference genome file - the empty spaces. Try to put only ">Chromosome1" (if you are working in fasta format) or delete the empty spaces of the header. Create again the bowtie index. Normally it solves the problem.

[Ajayi Oyeyemi]

tophat alignment

Hi everyone,
I have multiple fastq files and I want to align them all at once to tophat. Is there any shortcut that could be adopted instead of mapping each reads one by one to tophat.

Thanks.