This page contains a list of ideas to be implemented when I will find some time to do it :-(
Date wish emitted |
Program |
Description |
2005/04/09 |
compare-features |
- A program to compare two lists of features. Useful for evaluating
pattern matching results.
- The comparisons will be performed either at the nucleotide level or at
the site level (with a threshold on the mathcing nucleotides per pair
of sites).
- This program will compare a query and a reference set of positions and
return intersection, differences, calculate TP, FP, FN, TN, and
derived statistics.
|
2005/04/09 |
convert-markov-model |
- Read and write Markov models in different formats (oligo-analysis,
meme, Motifsampler).
|
2005/04/09 |
convert-matrix |
- Export all matrices when the input file contains several of them
(e.g. consensus, meme, gibbs).
- Calculate the significance sig=-log10(P-value) of each letter at
each position. The P-value would be calculated with a binomial model.
- Improve the criterion to include a letter in the consensus (use
significance rather than weight ?).
|
2005 |
dyad-analysis |
- The statistics for counts on 2 strands is still in the old mode
(single pattern probability, double sequence length). I should adapt
it in the same way as oligo-analysis.
|