I am just getting familiar with one of the many packaged that do this kind of prediction because I need to figure out the implemented scoring logic. The package uses three steps for targets selection. Namely: - sequence complementarity using a position-weighted local alignment algorithm - free energies of RNA-RNA duplexes - conservation of target sites in related genomes The energy minimization algorithm looks wierd to me. I don't quite get the other two phases. I was hoping something similar can be reproduced using Biomart functions to the purpose of understanding the package scoring criteria. After all, it looks like a sequence alignment problem ... Thank you, Maura -----Messaggio originale----- Da: Sean Davis [mailto:seandavi@gmail.com] Inviato: dom 09/08/2009 19.36 A: mauede@alice.it Cc: Bioconductor List Oggetto: Re: [BioC] Pairwise Sequence Alignments On Sun, Aug 9, 2009 at 10:15 AM, <mauede@alice.it> wrote: > I am browsing through the Biostring on-line docu,mentation. I wonder > whether the illustarted techniques (functions) for Pairwise Sequence > Alignments are being usd in any of the packages aimed at predicting miRNA > target-genes match. > I only haveonly used miRANDA which makes us of the Smith-Waterman > algorithm. However I cannot recognize any of Biostring techniques for > sequence alignment. I may be confused by the lack of documentation & > comments in miRANDA C code. > Thank you in advance, > Hi, Maura. This not an answer to your question, but miRNA target prediction is not just a function of pairwise alignment and is a research field unto itself (and not a field I am closely aligned with), so your best bet is to use published algorithms and predictions unless you have a specific need to improve on existing techniques (which implies understanding every detail of what they do). Sean tutti i telefonini TIM! [[alternative HTML version deleted]]