Tweet
Dear all,
I'm working with RNA-seq paired-end data and I'm using TopHat-BowTie in order to align my reads against the Human reference genome.
My problem is the following: I' have some reads coming from genes that are duplicated in the genome and they are very similar in terms of sequence, the only difference between them is that the second gene doesn't have the introns (for example this is the case of GLUD and GLUD2, or NACA and NACA2, etc..). When TopHat works with these reads it prefer align with mismatches in the contiguous sequence (GLUD2 or NACA2) instead of splitting the read in two parts and align it in two separated exon without any mismatch (GLUD or NACA).
Obviously the correct alignment is the second one, I'm sure of this because I got the same mismatches in all the samples that I analyzed.
How can I overcome this problem?
Can I use some option or particular parameter? I would like to constrain TopHat to prefer no-mismatch alignments even if they are splitted in two of more segments.
I hope my question is clear
Many thanks!
I'm working with RNA-seq paired-end data and I'm using TopHat-BowTie in order to align my reads against the Human reference genome.
My problem is the following: I' have some reads coming from genes that are duplicated in the genome and they are very similar in terms of sequence, the only difference between them is that the second gene doesn't have the introns (for example this is the case of GLUD and GLUD2, or NACA and NACA2, etc..). When TopHat works with these reads it prefer align with mismatches in the contiguous sequence (GLUD2 or NACA2) instead of splitting the read in two parts and align it in two separated exon without any mismatch (GLUD or NACA).
Obviously the correct alignment is the second one, I'm sure of this because I got the same mismatches in all the samples that I analyzed.
How can I overcome this problem?
Can I use some option or particular parameter? I would like to constrain TopHat to prefer no-mismatch alignments even if they are splitted in two of more segments.
I hope my question is clear
Many thanks!