Hi all. Got this from Damian Kao - a former visitor to the blaxter lab (back when he was a Postdoc in Oxford Dept of Zoology) who is also on the code monkeys group. Am forwarding on his behalf:
I am e-mailing because I recently just started working for a pet genomics startup. We are currently looking for someone to do a quick one-off job.
We are working with cats for now which has a pretty decent genome. Most recent genome version came out late last year and has defined chromosomal scaffolds (pacbio hybrid assembly). The goal of this job is to use low coverage samples (0.2-0.4X) in conjunction with the publicly available WGS data to impute and assign breed probability to the low coverage samples.
A more technical description of the job:
Someone who can take an existing .vcf of ~70 individuals (20-40X WGS) as a reference panel with defined breeds (breed replication is not very high). Then using the reference panel, impute a set of very low coverage WGS data (0.2-0.4X coverage). And finally use these imputed variants to try to assign a breed probability.
I will provide the reference .vcf and a set of 10 low coverage samples. The work is mainly analysis, preferably done in a jupyter/ipython notebook/R-studio using open source software (Beagle, admixture, scikit-allele, whatever). We will then take this analysis and turn it into a pipeline, so the analysis will need to be very well documented.
I would prefer an expert to just do the analysis. But I am also open to an expert who will be able to answer technical/scientific questions about the entire process. I already have a lot of the analysis done actually, but since this is not my specialty, I would like different opinions.
If you are interested or know people/grad students/post-docs that can do this kind of work relatively quick, please let me know.
----
If you are interested, or know anyone who is, please / ask them to contact
Damia...@gmail.com directly
- Sujai