On 10/25/2011 07:05 AM, wang peter wrote: > dear Martin: > your answer is quite clear. i will > try Biostrings::matchPDict to map the reads of each sample to the > assembled transcriptome using those samples. > > but what i worrry is the efficiency. i can also use BWA OR > bowtie to do so. > > i donot know if Biostrings::matchPDict will be very slow? matchPDict would not be my first choice; it will consume a lot of memory and will not be as flexible as some aligners. Martin > > thank you > > shan gao > -- Computational Biology Fred Hutchinson Cancer Research Center 1100 Fairview Ave. N. PO Box 19024 Seattle, WA 98109 Location: M1-B861 Telephone: 206 667-2793

Hi Shan Gao, Neither BWA or Bowtie can map junction reads which span two or more exons. You may try the Subread aligner implemented in Rsubread package which can map junction reads in addition to exonic reads. The percentage of junction reads in an RNA-seq dataset is typically around 20%. So you should be able to map ~20% more reads using Subread compared to Bwa or Bowtie. The Rsubread package also includes a function called featureCounts which counts the number of reads falling into each gene or each exon and returns you a list object which contains a table of read counts and also annotation information. If your worry about the efficiency, this might be the package for you. Rsubread is at least twice as fast as bowtie (and a lot faster than Bwa). The featureCounts() function only takes ~2 minutes to summarize mapping information from a SAM format file into a table of read counts. Cheers, Wei On Oct 26, 2011, at 1:05 AM, wang peter wrote: >> >> dear Martin: >> your answer is quite clear. i will try Biostrings::matchPDict to >> map the reads of each sample to the assembled transcriptome using those >> samples. >> >> but what i worrry is the efficiency. i can also use BWA OR bowtie >> to do so. >> >> i donot know if Biostrings::matchPDict will be very slow? >> >> thank you >> >> shan gao >> > > [[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor ______________________________________________________________________ The information in this email is confidential and intend...{{dropped:6}}

Hi, Shangao. You should demonstrate that you have followed up on previous replies and have read the relevant documentation before asking the list to provide answers. In particular, please read the documentation for scanBam and readGappedAlignment as Martin has suggested. You should TRY these functions on your BAM files. A 10GB BAM file is not large, so with a little reading, you may find that folks have thought about your perceived problem ("large" BAM files) rather carefully. I hope my comments above are taken as constructive; they are meant that way and not meant to be rude. Sean On Tue, Oct 25, 2011 at 12:09 PM, wang peter <wng.peter at="" gmail.com=""> wrote: > thank u. > but the bam file is still very large, more than 10G > i think i must read them partly by R. > so any good idea? > > shangao > > ? ? ? ?[[alternative HTML version deleted]] > > _______________________________________________ > Bioconductor mailing list > Bioconductor at r-project.org > https://stat.ethz.ch/mailman/listinfo/bioconductor > Search the archives: http://news.gmane.org/gmane.science.biology.informatics.conductor >