I briefly looked at the paper. Quite nice. I would probably go for that if I had constructed my own mask.
There are a number of mappability tracks in the UCSC browser many of which have different/similar k-mer versions. I choose the ones that are similar to the k-mer lengths we use in our mapping process but my question is: What are the main issues when deciding which one of these tracks to use? Are they simply interchangeable if they have the same k-mer option? What is your heuristic to choose from among those tracks when you use them?
I would like to see if there is a relationship between the mappability of a region (as defined by the track) and the genotypes I detect using exome sequencing (illumina)
Thanks
1 answer
You might want to look at this paper: http://www.plosone.org/article/info%3Adoi%2F10.1371%2Fjournal.pone.0030377
I know that there is a group (different from the one in the paper above) who are developing a general tool for generating mappability statistics for your particular genome and reads. I will update here when I hear that it has been released.
Is this not the GEM mappability track that is available inthe UCSC genome browser? If so I guess the answer is to use the GEM track as opposed to broad, duke or Umass tracks.
Log in to answer this question.
Can you give more information on what you're using these mapability tracks for?
I would like to see if there is a relationship between the mappability of a region (as defined by the track) and the genotypes I detect using exome sequencing (illumina)
Almost certainly there is a relationship.
Almost certainly there is a relationship. Actually you do not need a track. The mapper already tells you which region tend to be repetitive.