andreasdahl / genomeclustering Goto Github PK

Section about specific optimizations for our algorithm

When using KMers, a certain amount of negating errors is almost guarantied. Find the percentage of errors, a find a way to eliminate them as well as possible.

Try to use a Least Recently Used cache for clusters based on hits

Sorting by k-mers may give us a better cluster hit rate.

I get an out_of_range exception at https://github.com/AndreasDahl/GenomeClustering/blob/master/src/MufDifference.cpp#L146-L148 in most cases

Short an precise description of the problem

Upon inspecting the code it seems like FastaIO getNextLine at maximum reads two non ">" lines at a time and returns them in the string.

Change CMakeFile.txt to make it work with all needed compilers.

General section about existing clustering algorithms.

Currently all problem simply print the help text

We like to compare our distance measurements against the Levenshtein distance, but the Levenshtein distance takes very long to calculate, so big data sets takes ages to compare.

K-mer comparisons should be refined according to findings in #7