Hello Shuying, It might be best to contact the program BOWTIE's author, Cole Trapnell, for advice.
http://www.cs.umd.edu/~cole/research.html http://bowtie-bio.sourceforge.net/index.shtml "Mappability" is related to the methods used to map and what exactly you are mapping. Using the statistics from the genome sequencing effort is probably not going to give the best results for ChIP-seq data. There are some tracks in hg18 from the ENCODE project that predict mapability for various types/lengths of query data. Look under the track group "Mapping and Sequencing" with super track name = Mapability. The methods used for the different sub-tracks should be reviewed to see if any are appropriate. This wouldn't be "one parameter" information as-is, but how to use (summarize, window, etc.) this type of information is one example where the program's author might be able to offer guidance. We hope this helps, Jennifer --------------------------------- Jennifer Jackson UCSC Genome Bioinformatics Group http://genome.ucsc.edu/ On 3/1/10 6:35 AM, S. Sun wrote: > Hi all, > > I am trying to use BOWTIE to do alignment for chipseq data. I need to > provide one parameter named "fraction of human genome mappable by > reads". The default is 80%. Is this a good default parameter for hg18? > > In addition, I found the genome size statistics from the following web: > http://genomewiki.ucsc.edu/index.php/Genome_size_statistics > > It says that, for hg18, the total size is 3,107,677,273 and called > bases in the assembly is 2,881,421,696. Can I do the following simple > calculation: > 2,881,421,696/3,107,677,273 = 92.7%, and claim that "fraction of > human genome mappable by reads = 92.7%"? > > Thanks for your attention. I will be looking forward to your kind reply. > > Best wishes, > > Shuying > _______________________________________________ > Genome maillist - [email protected] > https://lists.soe.ucsc.edu/mailman/listinfo/genome _______________________________________________ Genome maillist - [email protected] https://lists.soe.ucsc.edu/mailman/listinfo/genome
