Question About Analyse Encode Data From Ucsc

Question

[e.karasmani]

Dear All,

I have download from the following link http://hgdownload.cse.ucsc.edu/goldenPath/mm9/encodeDCC/wgEncodeSydhTfbs/ some chip-seq data in order to do some analysis

so i am using a script that i know it works in order to convert the bam to bed (i know it works since i have done that iin the past many times )

my problem is that i get the following message

 Error: UserArgumentMismatch
 matrix requires identical quality score widths

 or the following errror
   Error in .local(x, width, fix, use.names, ...) :
   'x' is not a multiple of 'fix' length

depending on which encode chip i analyse

could you please give me ideas of what these errors mean, and how should i fix them?

I guess it is something about the bam file.....

thank you in advance

best regards Eleni

[Niek De Klein]

what script are you using?

[e.karasmani]

hi, i am using a script from the lab which uses ShortRead package and IRange to convert bam to bed

BUT i know that this script is working since i have used it million times in the past

so i guess the error that i get has to do with the bam from UCSC...

do you have any idea what these errors might mean for the bam?

Answer 1

[Istvan Albert]

It is hard (impossible?) to fix an error in a third party script that is not publicly available - also wether or not it worked in the past is also not indicative of a data error, when it comes to BAM format versus some home grown script it still a lot more likely that the script is flawed in ways that only now became apparent.

That being said iI would try to convert BAM to BED with other methods, say bamToBed from bedtools, see if that works or not, that will take you a step closer.