When Using Cuffdif Can I Combine Bam Files Obtained From Different Experiments?

Question

[narges]

Hi all,

I wanted to ask whether the order of genes in two different BAM files when running cuffdiff is important or not? To be more precise:

I have two BAM files which are the result of applying tophat over two samples from different experiments but the same reference genome (hg19).

Do you think that since they are from different experiments and the order of related genes may vary in these BAM files, I can obtain a meaningful cuffdiff result from them or not? Thank you in advance

[Assa Yeroslaviz]

what is the point of asking the same question twice? http://www.biostars.org/post/show/52699/cuffdiff-algorithm/

[narges]

I have not asked it twice, just somebody has improved the title of my question for me.

[Istvan Albert]

I changed the title to make it more informative - when this happens it may show up as new entry in a Feed reader if that buffers the content - it is a limitation of the feed reader.

[narges]

thank you for the help, now I understand I should use more informative title

Answer 1

[Damian Kao]

Cufflinks suite of software requires your bam/sam file to be sorted by reference position. Tophat generates a sorted file that is appropriate for Cufflinks/Cuffdiff. The order of the genes should be the same in your bam files.