The Best Tool for RNA Alignment

[EthoStrain]

Hello,

I have recently been trying to align both RNA and DNA reads to a reference sequence using BWA and bowtie. While they have been working wonderfully for the DNA, the RNA is being handled with hostility.
The reference and the reads both contain Uracil but it seems that both BWA and bowtie think that Uracil is lame and subsequently convert it into Thymine or 'N'. I was looking at tophat as a way to solve my problems but before I spend time configuring it I wanted to get some input from people on this forum.

Example:
I want to align this:
agcgaguaaaagacu

against this:

ccucuuuccaucuacagcccacuggcacuccaggcaccuaccuacuucaaacaagcucaa
gccaaggccuuccccuaacucugacugcuagucccacaguaacccugacagcugcugcuc
cugcuucuccugaacagauuauuguucaugcuuuauccccagaacauuuguugaacacaa
gugauaauguuacagugcagugucacacaccaagagucaucauucagacuguugccacag
aggacaucacuucuuccauaucccaagcagaacugaccgucgauagugauauucagucau
cugauuuuccugagccuccagacgcccuagaagcagacacuuucccagaugaaauucauc
acccuaagaugacuguggagccaucauuuaaugaugcucauguauccaaauucagugaccaaaauagcacagaacugaugaauaguguuauggucagaacagaagaagaaaucucugaca
ccgaccuuaaacaagaggaaucacccucugauuuagccagugcuuauguuacugaggguu
uagagucucccacuauagaagaacaaguugaucaaacaauugaugaugaaacaauacuua
ucguuccuucaccacauggcuuuauccaggcaucugauguuauagauacugaaucugucu
ugccuuugacaacacuaacagaucccauacuccaacaucaucaggaagaaucaaauauca
uuggaucauccuugggcaguccuguuucagaagauucaaaggaugucgaagauuugguaa
acugucauuagaauaauucu

What is the best tool to do this. Please keep in mind that I have a very large reference sequence so speed is critical - One of the main reasons I went with a BWT based tool to begin with.

Thank you!

[frozenlyse]

Where are you getting these sequencing "reads" from? Anyway just replace all your U's with T's

[lexa]

might be a stupid question but why don`t you replace all Us into Ts and then run bwa or bowtie? this can be done with an easy perl script or even an awk one-liner.

edit: there is probably not a single best tool available for your problem.

[NGSfan]

Just replace the U's with T's.

cat myRNAfasta.fa | sed -e 's/u/t/g' > myDNAfasta.fa

note that this will change all lowercase 'u' to 't' , even in your fasta header!

[EthoStrain]

Not quite what I need

Hi all,
I have already arrived at the conclusion that everyone has come to, but I am looking for an alternative.
Lets just assume that your reference has both DNA and RNA. I suppose this isn't going to happen in 'normal' circumstances so it won't be supported? I was certainly surprised when these aligners didn't allow for Uracil, but I can understand the reasoning.

I thank you for your replies!

[Thomas Doktor]

With the current sequencers, you are always sequencing DNA which might be a cDNA made from an mRNA. How did you end up with sequences with U in them? And why not simply align them to a DNA reference after conversion?

Btw, TopHat uses Bowtie to align the sequences so whatever restrictions Bowtie imposes on the reads and the reference will be the same for TopHat.