methylation ratio Error

[atongo77]

Hello folks,

I was using the script "methratio.py" from BSMAP to extract methylation frequencies on reads from bisulphite sequencing. I have been getting this error message on some specific chromosomes. I am using a reference genome of mm9.fa.


Traceback (most recent call last):
File "methratio.py", line 253, in <module>
disp('total %d valid mappings, %d covered cytosines, average coverage: %.2f fold.' % (nmap, nc, float(nd)/nc))
ZeroDivisionError: float division by zero

I would appreciate it if someone can help me figure out what might be going on.

Thanks.

[dpryan]

Can you show the exact command used (and perhaps the first 2 or so alignments in your BAM file)? This should only ever happen if you have no covered cytosines, which would seem rather unlikely.

BTW, keep in mind that there are innate problems with using BSMAP's output, since you can't filter by quality. Having said that, since bsmap is much faster than anything else out there I can't really fault you for using it

[atongo77]

@dpryan

Hello dpryan,

Thank you very much for your response. This is the command I issued:

> python methratio.py --ref=mm9.fa --out=brain_control_chr9.txt --chr=chr9 --context=CG --zero-meth *.bam

The "bam" files were about 5 in all. also used the "samtools" to issue this command on one of the bm files:

>samtools view 2_SRR955122*.bam | head

And this is what I got:

SRR955121-NP070.2 0 chr2 263592509 255 36M * 0 0 CTCTGCCTGCCTCCCCTCTTGNGAATCTTGCAGGCT ggggggggggggggggceggdGddda_a__gegfgd NM:i:0 ZS:Z:++
SRR955121-NP070.3 0 chr14 60374539 255 36M * 0 0 GGAAGGACGCGCACACCTGCANTCGTATGTGTACAC ffggggdggggggggfggcfbF^aaSSVVSdfdffa NM:i:0 ZS:Z:++
SRR955121-NP070.5 16 chr1 187870119 255 36M * 0 0 TGTGAATTCAGTTANTTCCCAGTAGCCTATTCTTGT ggcgggaaaa^aa^Eagagggggggggdcggggggg NM:i:1 ZS:Z:-+
SRR955121-NP070.6 16 chr11 35334197 255 36M * 0 0 CATGGAGGCGATAAGCGATCCGAACATCAGCATGAA ffcfd]ffcaddaffffdfffffaffffafcfddfd NM:i:0 ZS:Z:-+
SRR955121-NP070.13 16 chr2 20010899 255 36M * 0 0 TCGTAAGAACTCACGCTCTTCTAAGAAAGGGAGGTA dgggcgffffffffffedcggfffdfagggdggggg NM:i:0 ZS:Z:-+
SRR955121-NP070.16 0 chr2 226862985 255 36M * 0 0 TGTGGTGCAATTCCAGCGGTGATGACGGAACTGGTG eae`dfcdfc`aaaaf_cfff_acfff\fdeSce]f NM:i:0 ZS:Z:++
SRR955121-NP070.17 16 chr16 41229004 255 36M * 0 0 CCCATTGAGCATTTCTATGTAGGTAGTTTAAGGCCC a^S]aaYU\WNYYUYPL]JXU[U\QQcccc\cccc_ NM:i:0 ZS:Z:-+
SRR955121-NP070.18 0 chr15 36462077 255 36M * 0 0 CCAAGGTCCAGGGACAAACCATGGGAGAGGGAGTAA f]fffaaddda^Yacdaddd[dd\dUWIYVWT_SXf NM:i:0 ZS:Z:++
SRR955121-NP070.19 16 chr7 117682445 255 36M * 0 0 AGTAGCCCGCAAGTTTGGAAGTAGCGCTGGGCCGCT c[\O_I^ac[aZ\\X\b^Y\acY^_a^cccccaccc NM:i:0 ZS:Z:-+
SRR955121-NP070.22 0 chr1 187649295 255 36M * 0 0 AGCACTGTTTCTTTGAATTCAGTTAGTTCCCAGTAG gfgggggggggfgggfggggaggggggfgggggggf NM:i:0 ZS:Z:++


I really appreciate your help. In addition, I would greatly appreciate any advice on how to be efficient and effective in this analysis. Thanks.

[dpryan]

Can you confirm that you have alignments on chr9? If so then I haven't a clue what the issue is (I'd have to go through the source code). You might try to contact the authors in that case.

[atongo77]

Thanks dpryan. I think there are no alignments on chr9. It appears this problem crops up for all chr* for which there are no alignments. I have seen similar errors for chrX and chrY. I have sent a mail to the authors and waiting for their response. Thanks once again.

[atongo77]

Hello dpryan,

Were you able to check from the source code on what might be causing the division by zero error? Authors are yet to respond to my email. I truly appreciate your help.

Thanks