baezortega
commented
7 years ago TODO: Improve documentation of all functions, including "Details" and "Value" fields.
--
Outline of the vignettes: Example 1: fitting to simulated sample:
- TODO: Plot the signatures to use for the simulation, using plot_spectrum() and par(mfrow=...), before fitting.
- Add a comment saying that the "emu" method will be shown in the next example (without running it).
Example 2:
Do a proof of concept by fitting COSMIC signatures to a simulated sample. Show exposures and reconstructed spectrum.DoneLoad 21 breast cancers data (mutations_21bc). Use the build_catalogues() function to build the 21 breast cancer catalogues (and suggest ways of obtaining the trinucleotide contexts, e.g. BSGenome or the VCF INFO field).DonePlot 21 genomes in multiplot grid.Done- Fit signatures to data using "emu" instead of "nmf".
- Plot extracted exposures.
- Plot reconstructions (add comment about how to output to PDF).
- Extract a range of signatures (2:10) using "emu" with small sample size (include a comment showing "nmf"). Show GOF.
- Extract again for the estimated best number of signatures, with more samples.
- Plot extracted signatures and exposures.
- Plot reconstructions (add comment about how to output to PDF).
Extend vignettes, e.g. including multi-sample fitting and explaining the required dimensions and contents of the input matrices (now it’s quite ambiguous), as well as all the post-sampling functions (below). Add the browseVignettes function to the GitHub Readme.