Fixed bug in train.HMM causing log likelihood to vanish for Baum Welch training (thanks Marie-Amelie Forin-Wiart for bug report)
Enabled full array output for posterior.PHMM (thanks Leon Eyrich Jessen for bug report)
Fixed bug in align
not closing cluster after multithread
Fixed bug in align
preventing colnames for 2-row alignments
Enabled 'near-enough' alignment matching between Viterbi iterations in train.PHMM
via limit
parameter
Prevented memory overflow in 'train.PHMM' caused by md5 hash attmpt on very large alignments
Included faster maximum entropy sequence weighting option (Henikoff & Henikoff 1994)
Enabled multithread option for Baum Welch training of profile HMMs.
Fixed checksum parse issue in readPHMM
Added posterior probabilities for PHMMs.
derivePHMM.list seed selection changed from max length to most common length.
Fixed issue in 'align' causing distance matrix error if seqweights not set to NULL (Thanks to Ben Margetts for bug report).
Updated dependency list to include kmer instead of phylogram.
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.