malariagen / ag1000g-phase1-vgsc-report Goto Github PK
View Code? Open in Web Editor NEWLicense: MIT License
License: MIT License
Finalise recombination figure(s) for partner review.
Finalise text of functional variation sub-section ready for partner review.
Need to rebuild the variants table with correct phasing of multiallelics.
Fill in authors and affiliations.
Revisit the figure using haplotype age distributions within outbreaks to present evidence for direction of spread. Revisions for partner review.
Discussion paragraph on accessibility of the gene, have we missed any variation, etc.
Text for methods section sufficient for partner review.
Figure(s) showing results for assay design sub-section. Possibly (1) figure showing information gain per SNP over the region, and/or (2) figure showing cross-validation results for decision trees with increasing numbers of SNPs.
Revisit the map, make some tweaks to convey some additional uncertainties.
Finalise text of recombination section for partner review.
There are new releases of graphviz Python wrapper (0.6.0) and SciPy (0.19.0). Worth updating at some point.
Make any edits after revisions to results and discussion, e.g., soften language about predicting phenotype (e.g., say discuss putative phenotype).
The notebook Figure_1_RelativeAges.ipynb uses sample metadata from phase 2. If this is needed, load from external release directory rather than copy into data folder.
Review text of the "Insecticide resistance outbreaks" sub-section, any changes to suggest prior to partner review?
Hi @alimanfoo
Here's the error from when I tried to used locate_recombinants with the largest haplogroup. All other haplogroups worked fine.
IndexError Traceback (most recent call last)
in ()
1 #remove recombinants - F3 won't build if max-dist is >1 - maybe too complex
----> 2 idx_rec = hapclust.locate_recombinants(cluster_haps, debug=False)
3 #idx_rec
~/Git/agam-vgsc-report/agam-report-base/src/python/hapclust.py in locate_recombinants(h, debug)
816 def locate_recombinants(h, debug=False):
817 """Locate recombinant haplotypes via the four gamete test."""
--> 818 count = count_gametes(np.asarray(h, dtype='i1'))
819 d = np.all(count > 0, axis=(2, 3))
820 # indices of recombinant haplotypes - N.B., keep track of different possible solutions
~/Git/agam-vgsc-report/agam-report-base/src/python/hapclust_opt.pyx in hapclust_opt.count_gametes (/home/chris/.pyxbld/temp.linux-x86_64-3.5/pyrex/hapclust_opt.c:2346)()
IndexError: Out of bounds on buffer access (axis 3)
I think a label is missing for a red edge into the big node. No urgency to fix for partner review but raising so we don't forget.
Revisions to abstract and intro to get ready for partner review.
Finalise text of sub-section about design of genetic assays for outbreak surveillance.
Add numbers and possibly a supp fig to address question of whether EHH is significantly different between different core haplotypes.
E.g., compute integrated haplotype homozygosity (IHH) for each core haplotype, compute confidence intervals somehow, maybe plot with error bars.
When this code is finalised to produce the supplementary recombination figure, change the name of the notebook
Create a figure showing haplotype structure.
Change "predicted phenotype" column to "phenotype"
Change column text to "known/putative"
Collapse duplicate multi-allelic rows
Add aedes reference - evidence
Add the numbers of intronic and exonic SNPs used in the global analysis.
Make single discussion subsection on these topics.
E.g., DK mentioned sentence, "large nodes consistent with positive selection, suggestive of a functional role in pyrethroid resistance...". Our work raises questions about functionality that need a concerted effort to resolve experimentally.
Add reference to the paper on combination of substitutions in 1534 and 410 in Aedes.
In the assay design section we end with how great it is that we can evaluate these haplotypes in the field using just 20 SNPs. Make clear that this is possible because of the preceding genomics.
Check through text, any missing references needed before partner review?
Add number(s) quantifying the correspondence between the clusters obtained via the networks and the clusters previously obtained via hierarchical clustering.
Currently network code breaks if haplotypes with multiallelics are used in conjunction with network_method='mjn'
Being able to discern network cluster/node membership would be useful to check concordance with other methods.
Discussion paragraph on what are the implications for insecticide resistance management? Realistically how could this information be used?
Nick has found a bug in the AR3.1 release such that the Zarr format files for the haplotypes are not correct and are not the same as the HDF5 format files for the haplotypes.
Currently our analyses use the Zarr format files. We can work around this by using the HDF5 format files instead everywhere. This will require modifying the setup modules to use the HDF5 files, then rerunning all notebooks that depend on the haplotypes (table 1, figure 1, ... everything basically).
Discussion paragraph on weaknesses, caveats and potential improvements to method for estimating haplotype age.
Finalise the haplotype tree figure for partner review.
Finalise text of the haplotype age outbreak analysis sub-section for partner review.
Pick a journal to submit to, do minimal formatting required for initial submission.
Initial pass on methods, break down into further issues as required.
Finalise text of haplotype structure sub-section ready for partner review.
Finalise haplotype networks figure for partner review.
Possible discussion paragraph on why L995F but not L995S has enhancers.
Following on from work done in #94, there are still some unresolved points, including:
Davies et al 2007?
Finalise the figure with the overall haplotype age distribution for partner review.
Produce supplementary data tables as described in sub-section about assay design. Two tables envisaged: (1) table of SNPs with information to help primer design and initial SNP selection; (2) table with all haplotypes and alleles sufficient for building a classifier.
Discussion paragraph on DDT. If prior selection for DDT resistance, how might this complicate the picture? Do we see any evidence for multiple phases of selection?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.