malariagen / agam-kilifi-report-2017 Goto Github PK

Describe known resistance alleles at the Ace1 locus.

Placeholder for other analyses attempting to improve inference of recent Ne changes, particularly timing and magnitude of bottleneck.

Placeholder for work to look for genome regions under recent selection, esp. due to insecticides.

N.B., this probably not possible using conventional scanning metrics like h12 or IHS due to extreme demography, but maybe other things we could do, e.g., look at number of samples with ROH over the genome, or number of pairs with IBD, identify particularly homozygous/inbred regions?

Placeholder for work on any other resistance loci, other than those already specifically mentioned in other issues.

Compute and plot data on runs of homozygosity. Compare by sampling site.

Placeholder for work to elaborate on the initial AIM results showing Kilifi mosquitoes have a mixture of gambiae and coluzzii alleles.

Various hypotheses to test:

• Due to recent admixture with unknown coluzzii population East of Rift
• Due to ancient admixture with coluzzii-like population
• Due to ancestral polymorphism (Kilifi is an outgroup of gambiae/coluzzii)
• Due to admixture with some other species (arabiensis/merus/...)
• ...

Describe known or putative resistance genotypes at the Gste locus.

Just to note I'm having an install issue:

ERROR conda.core.link:_execute_actions(337): An error occurred while installing package 'conda-forge::graphviz-2.38.0-0'.
PaddingError: Placeholder of length '80' too short in package /github/alimanfoo/agam-kilifi-report-2017/dependencies/miniconda/envs/agam-kilifi-report-2017/bin/graphml2gv.
The package must be rebuilt with conda-build > 2.0.

From googling it looks like this is due to the install path being too long - using a shorter path would resolve. I may tweak the install script to try and shorten some directory names.

Run a PCA analysis on just the Kenyans, look for evidence of any population structure related to sampling location.

Compute IBD tracts between all pairs of individuals.

Plot summaries of IBD data (total IBD, no. tracts), comparing within and between sampling sites.

There seems to be a clear drop in TdD around the centromere of chromosome 3 in one population?
I'm not sure what may be causing this- is this a point for further discussion?

Anything we can add about recent Ne using IBDNe?

Split by sampling location?

Any way to get better uncertainty estimates? Concatenate and run both arms of Chromosome 3 together? Rerun multiple times jackknifing over samples?

Compute summary statistics of nucleotide diversity, including pi, watterson's theta, tajima' d, and the full site frequency spectrum.

Compute for whole population and for each sampling site separately. Any significant difference between sampling sites?

Describe resistance genotypes at the Cyp6p locus.

What should the report title be?

Characterise known or putative resistance genotypes in the Vgsc gene.

Use e.g. D statistic to scan for regions of the genome with evidence of adaptive introgression from arabiensis or other species.

Attempt to retrieve rainfall data for the local area, any correlation with our inferences regarding timing of bottleneck?

Consider trying to construct a network of individuals according to IBD sharing, to look for any fine geographical population structure.

Characterise the karyotypes present in the population at major inversions on chromosome 2.

Compute an LD decay curve.

For all Kenyan mosquitoes, and separately for each sampling site?