Chris Miller

Calculating mapability is really straightforward: Take a reference genome's fasta file and generate all possible reads of a given length. Map those reads back to the reference genome, and identify which of them map uniquely in the expected location. You'll find some very simple bash/perl code that wraps BWA and then calculates mapability and gc-content scores in 100bp windows here: https://code.google.com/p... Keep in mind that this code only generates single-end reads, but by using paired-end reads, you up your mapability significantly. The code could easily be tweaked to do that. Be sure to think about what insert size is appropriate. There are also some precomputed mappability tracks available for download through the UCSC table browser. - Chris Miller

The visualizations that will be most useful depend to a huge extent upon the design of the study. There are many, many things you might want to explore in this data, so you're going to have to narrow it down to get reasonable recommendations. - Chris Miller

This would probably be better off posted as a new question. It's likely that only the two or three people involved in this thread will notice that you've posted it here. That said, This previous question may help: What is the default quality encoding expected by BWA? - Chris Miller

What you're asking here is probably beyond the scope of a Q/A site. To properly review all of these steps and provide feedback and suggestions would take hours. If you really need that level of support, then you're going to want to pay someone a consulting fee to help you get your pipeline set up. If you have specific questions about individual steps or commands, then Biostar can be a great resource, and please do feel free to ask questions. I'd encourage you to look through old posts first, as many of these topics have been addressed individually in the past. - Chris Miller

The squaredancer code is buried in one of the other repos. Here's a direct link to the perl script: https://github.com/genome... - Chris Miller

The first result for a search on "ncbi_enc" is this page: http://www.ncbi.nlm.nih.gov/books... It says: "The data files distributed through the dbGaP are all encrypted by NCBI’s special encryption algorithm. These files have a file suffix “.ncbi_enc”, indicating that they are NCBI encrypted files." That page also contains a link to the archive and encryption utilities. - Chris Miller

That doesn't really make sense. "Low-pass" generally refers to a genome that's sequenced to a depth under 10x. With this data, you can call germline SNPs, find structural variants, etc. It's not particularly useful for cancer sequencing though, as somatic variants are difficult to discern and forget about finding subclonal variants. - Chris Miller

This is a bad idea. Since the genome assembly that the reads were mapped to are different, you really need to realign your data. There will undoubtedly be many places where reads map to different places than where liftover would place them, due to the differences between the assemblies. Convert the bam back to a fastq with picard, then redo the mapping with the aligner of your choice. - Chris Miller

Yes, your comment is way off topic, but I'll briefly respond to say that I haven't seen this behavior. In grad school, I was in a small lab with no direct connection to TCGA and we had no problems getting access to the protected data. Yes, you need to state a rough research plan so that they can verify that you'll safeguard protected patient information. Yes, you also need to wait until the marker paper is published, as the people who worked so hard to generate the data get the first shot at one general paper describing the dataset. I don't feel like that's unreasonable. - Chris Miller

FWIW, Stack overflow is revamping how they deal with closing posts: http://blog.stackoverflow.com/2013... - Chris Miller

FWIW, MAF is a "standard" format within the TCGA project. Here's documentation: https://wiki.nci.nih.gov/display... - Chris Miller

If things are truly being passed strictly through pipes, this approach seems horribly susceptible to failures. If your plotting step at the end of a long sequence of steps fails, do you have to remap your entire WGS experiment? - Chris Miller

If you want to get a good publication "just on computational basis", you're doing it wrong. An algorithm can be interesting or useful, but bioinformatics is about making biological discoveries. If you haven't applied your tool and come up with some new and fascinating insights into biology, then no, you can't expect to publish in a selective journal. - Chris Miller

Can we get an option to link the Ad to the job posting? Seems like this would be one of the most common uses of ads for academic users. - Chris Miller

If you're referring to the ideogram at the top, then I believe you're overthinking it. There's a lot of heterochromatin near the centromere that isn't necessarily actually the centromere. - Chris Miller