Sign up ×
Biology Stack Exchange is a question and answer site for biology researchers, academics, and students. It's 100% free, no registration required.

One way to assemble fragments produced by DNA sequencing (often called reads) is to seek for the shortest common superstring that contains all the reads of a given set of reads. One model for this problem is the reconstruction model, which calculates the minimal edit distance between a possible superstring (that contains all fragments) and each one of the fragments. What I don't understand yet is the fact that when calculating the edit distance for a given fragment, it's necessary to calculate the edit distance for the reverse complement of that fragment.
An example of a fragment and its reverse complement could be:

ACTGTCC fragment
TGACAGG complement (the complementary strand as we know)
GGACAGT reverse complement

According to the book I'm reading the ideia behind it is to find a superstring as short as possible that either the given fragment or its reverse complement must be an approximate substring of the superstring. I know that many mathematical/computational models don't have a biological sense for their way of soluting a biological problem, but this recontruction model I could think it's biologically reasonable to find a string that contains all the fragments of a certain region of a DNA sequence, and all the complement fragments, after all when assembling we can find both strand sequence. But a reverse complement is not something that exist for a given sequence, unless it has suffered some kind of break-recombination event, which is something not expected, mainly when talking about conserved regions. Do you any idea why this model takes into account the reverse complements ? Thanks in advance.

share|improve this question
Isn't this just a case of keeping everything straight with respect to 5'>3' direction? If a read matched "TGACAGG" this would not correspond to either strand of your example, because writing it like that implies that it is 5'-TGACAGG-3' unless it is made explicitly clear that it is the bottom strand of a duplex. – Alan Boyd May 21 '13 at 5:24
So you're saying that in fact the model see every fragment in 5'->3' direction? It doesn't consider the real existence of a reverse complement, but simply uses it because the reverse complement is a way of seeing the complementary strand of a given fragment with 5'3'direction? I'm sorry about the slow reasoning. – Heidi Garcia May 21 '13 at 14:36
Yes, if you have a read (which will be by convention 5'>3') then you can search for it directly on the top strand of the target (also 5'>3'). If you need to look for it on the other strand of the target you can't simply search on the complementary because this is 3'>5', so you derive the reverse complement. It's just the equivalent of turning the dsDNA sequence (on the page) through 180 degrees. – Alan Boyd May 21 '13 at 15:52
Nice! Thank you very much, Alan. You helped me a lot. – Heidi Garcia May 21 '13 at 18:55

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.