hildebra/lotus2: LotuS2 2.21

-fixed critical bug in dada2 clustering: when using "SequencingRun" column that was unsorted, recruited reads could be significantly lowered -sdm v2.02: refactor code to make future work easier; slight speed increases -new option "-mergePreClusterReads <1/0>" to merge reads before sequencing clustering. Not recommended currently, cluster quality seems to be lower when merging reads. -various improved help texts