libcdhit

cd-hit library for yclust

TODO

• add source from yclust
• building scripts using cmake
• replace the full k-mer filter with weighted minhash
• memory efficient data struct for countWords function
• better multi-threading performance
• add APIs start from raw seqs and minhash sketches
• effcient SW, such as SSW or cuda enabled CUDASW++
• NUMA-aware for word table
• inverted index vs. direct pair-wise comparison for small datasets