crossSetSim() now gains two new arguments batches and verbose.
The batches argument allows users to split the similarity computations into multiple batches, which is useful when dealing with a large number of sequences and limited RAM. The verbose argument enables progress updates during the computation. This brings crossSetSim() to feature parity with parSeqSim(). (thanks, @ofleitas, #41)
A new function crossSetSimDisk() has been implemented as a disk-based version of crossSetSim().
This function follows a similar approach to parSeqSimDisk(), where partial results from each batch are cached on the hard drive and merged at the end. This allows for processing larger protein sequence sets that may not fit into RAM (#41).
crossSetSim() is added for calculating pairwise similarity between two sets of protein sequence based on sequence alignment (thanks, @seb-mueller, #34).extractProtFP() and extractProtFPGap() when index = NULL (thanks, @fcampelo, #30).system.file() usage to avoid confusion (thanks, @jonalv, #31).batches to parSeqSim(). The new argument supports breaking down the pairwise similarity computation into smaller batches. This is useful when you have a large number of protein sequences, enough number of CPU cores, but not enough RAM to compute and hold all the pairwise similarities in a single batch. Also, use the other new argument verbose to track the computation progress.parSeqSimDisk(). Compared to the in-memory version parSeqSim(), this new function caches the partial results in each batch to the hard drive and merges the results together in the end. This could further reduce the memory usage for parallel similarity computations involving a large number of protein sequences.parGOSim() that will create minor numerical inconsistencies in results due to argument matching.twoGOSim() and parGOSim() to use the latest GOSemSim API for computing GO based semantic similarity. Issues in the code examples are also fixed. We thank Denisa Duma for the feedback.getUniProt().gap.opening and gap.extension to parSeqSim(), allowing more flexible tuning of the sequence alignment for more types of amino acid sequence data. We thank Dr. Maisa Pinheiro for the feedback.removeGaps() for removing/replacing gaps (-) or any irregular characters from protein sequences, to make them suitable for feature extraction or sequence alignment based similarity computation. We thank Dr. Maisa Pinheiro for the feedback.ifelse conditioning (3f6e106) for the distribution descriptor in CTD. We thank Jielu Yan from the University of Macau for kindly reporting this issue.curl -I -L:
2015-12-28
extractCTDD().2014-12-18
2014-09-20
2014-06-20
LICENSE file according to CRAN policies.readFASTA().getUniProt().protcheck().protseg().