In a similar vein as #63 , the k-mer logic currently doesn't check neighbors, which drops reads that have a mismatch. Probably best to have a queue of positions to evaluate. Place the exact matches first (on either strand), check those to set a bench mark of best. Then check 1 neighbors of those positions.
In a similar vein as #63 , the
k
-mer logic currently doesn't check neighbors, which drops reads that have a mismatch. Probably best to have a queue of positions to evaluate. Place the exact matches first (on either strand), check those to set a bench mark of best. Then check 1 neighbors of those positions.