On 09/10/13 04:37 AM, 陈嘉鹏 wrote: > Dear Boisvert: Hello Dr. Lee,
> > I read the material and method section of your paper, but didn't find how you > concatenate contigs to scaffolds. In http://genomebiology.com/2012/13/12/R122 , we described the workflows "Ray Meta" (scalable metagenomic assembly) and "Ray Communities" (scalable kmer-based taxonomic and gene ontology profiling). We never really published anything about the scaffolder, but it is basically a greedy algorithm that uses paired information. > One possibility I think is using paired-end information >to cat two contigs if they have PE reads supporting they come from one DNA >fragment. However, in our case, sequencing insert size is just 180bp, but >having some >scaffold gaps more than 150 (Tablet visualization shown below). So this is >confusing. Yes, it uses paired sequences. If you look at LibraryStatistics.txt and LibraryData.xml, you will see the distribution of outer distances for your paired sequences.You probably had fragments that were longer than 180 bp and this will be reflected in these 2 files. The file RayOutput/ScaffoldLinks.txt contains the actual links detected and used by the scaffolding algorithm. Another thing you can do is compare the average coverage depth of the left contig (on your picture) and the average coverage depth of the right contig (on your picture). Since they are together in a scaffold, their coverage should be fairly similar. > 内嵌图片 1 > > What kind of miss-scaffolding, if there any, do you think Ray Meta may > produce? Usually, we have observed that the Ray scaffolder is not overzealous and does not overjoin contigs. For instance, see the Assemblathon 2 paper : http://www.gigasciencejournal.com/content/2/1/10/abstract > Miss-scaffolding may influence downstream analysis. Eg. we are gonna classify >the scaffolds to different bins based on tetranucleotide frequency patterns. You could also use the coverage depth on another axis. In Ray, this is available in the output coverage depths in RayOutput/BiologicalAbundances/_DeNovoAssembly/ (through the default "Ray Meta" workflow) . Coverage binning is a popular approach, see (behind a paywall): Genome sequences of rare, uncultured bacteria obtained by differential coverage binning of multiple metagenomes http://www.nature.com/nbt/journal/v31/n6/full/nbt.2579.html > But if contigs from different bacterials are concatenate in a scaffold, the > binning will fail. > > > 2013/9/5 Sébastien Boisvert <sebastien.boisver...@ulaval.ca > <mailto:sebastien.boisver...@ulaval.ca>> > > On 05/09/13 07:53 AM, 陈嘉鹏 wrote: > > Dear Boisvert: > > > > I found ray use same command for single genome assembly and > metagenomics assembly. But how does ray know the data is metagenomics or > single genome sample? > > > > Hello, > > Ray 2 (2.0.0, 2.1.0, 2.2.0) is an improvement over the older Ray 1 in the > sense > that the algorithms are generalized to work on both genomes and > metagenomes. > > In our paper http://genomebiology.com/2012/13/12/R122 , we explained how > we did it. > You can read more about that in the Materials and methods. > > Ray Meta is really just a workflow for metagenome assembly provided by > the Ray 2.x.y series. > > Also, Ray Communities allows you to profile gene ontologies and taxons in > your sample. > This is also part of Ray 2. > > > > > > 2013/9/5 陈嘉鹏 <kevinc...@gmail.com <mailto:kevinc...@gmail.com> > <mailto:kevinc...@gmail.com <mailto:kevinc...@gmail.com>>> > > > > Our cluster administrator says: Seems it is caused by the library > file libdl.a. It should be fixed now. > > This is the only thing I know. > > > > > > 2013/9/4 Sébastien Boisvert <sebastien.boisver...@ulaval.ca > <mailto:sebastien.boisver...@ulaval.ca> > <mailto:sebastien.boisver...@ulaval.ca > <mailto:sebastien.boisver...@ulaval.ca>>> > > > > On 04/09/13 05:22 AM, 陈嘉鹏 wrote: > > > Dear author: > > > We have solved this problem, but still have other problems. > > > But you don't need to think about my last email. > > > > Hello, > > > > Just out of curiosity, what was the problem during the > compilation of Ray > > that produced " undefined reference to `__dlopen'" ? > > > > And how did you solve it ? > > > > > > > > Thanks > > > > > > > > > > > > 2013/9/4 陈嘉鹏 <kevinc...@gmail.com > <mailto:kevinc...@gmail.com> <mailto:kevinc...@gmail.com > <mailto:kevinc...@gmail.com>> <mailto:kevinc...@gmail.com > <mailto:kevinc...@gmail.com> <mailto:kevinc...@gmail.com > <mailto:kevinc...@gmail.com>>>> > > > > > > Dear Boisvert: > > > > > > Thanks very much for developing MPI meta assembler. Our > cluster has only this choice. I used below command to compile: > > > make > PREFIX=/disk/rdisk08/jiapchen/bin/Ray-v2.2.0/ray-build > > > > > > But I got error message below: > > > > > > > /share/apps/mvapich2-1.6/lib/libibverbs.a(src_libibverbs_la-init.o): In > function `load_driver': > > > (.text+0xc7): warning: Using 'dlopen' in statically > linked applications requires at runtime the shared libraries from the glibc > version used for linking > > > /share/apps/mvapich2-1.6/lib/libdl.a(dlopen.o): In > function `dlopen': > > > (.text+0x5): undefined reference to `__dlopen' > > > /share/apps/mvapich2-1.6/lib/libdl.a(dlclose.o): In > function `dlclose': > > > (.text+0x1): undefined reference to `__dlclose' > > > /share/apps/mvapich2-1.6/lib/libdl.a(dlsym.o): In > function `dlsym': > > > (.text+0x5): undefined reference to `__dlsym' > > > /share/apps/mvapich2-1.6/lib/libdl.a(dlerror.o): In > function `dlerror': > > > (.text+0x1): undefined reference to `__dlerror' > > > collect2: ld returned 1 exit status > > > make: *** [Ray] Error 1 > > > > > > I know the reason may be in our cluster, but I don't > have hint after some google search. So could you please give me any ideas? > > > > > > Thank you very much, > > > Jiapeng > > > RA in City University of Hong Kong > > > > > > > > > > > > > > > -- > > > Jiapeng Chen, Research Assistant > > > # School of Energy and Environment > > > # City University of Hong Kong > > > Shatin, N.T. HK China > > > Cell phone: +852-4789219 > > > > > > > > > > -- > > Jiapeng Chen, Research Assistant > > # School of Energy and Environment > > # City University of Hong Kong > > Shatin, N.T. HK China > > Cell phone: +852-4789219 > > > > > > > > > > -- > > Jiapeng Chen, Research Assistant > > # School of Energy and Environment > > # City University of Hong Kong > > Shatin, N.T. HK China > > Cell phone: +852-4789219 > > > > > -- > Jiapeng Chen, Research Assistant > # School of Energy and Environment > # City University of Hong Kong > Shatin, N.T. HK China > Cell phone: +852-4789219 ------------------------------------------------------------------------------ October Webinars: Code for Performance Free Intel webinars can help you accelerate application performance. Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from the latest Intel processors and coprocessors. See abstracts and register > http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk _______________________________________________ Denovoassembler-users mailing list Denovoassembler-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/denovoassembler-users
