fastLSA is a new implementation of the Local Similarity Analysis (LSA) algorithm that approximates the detection of significant LSA results using an asymptotic upper bound. Making only the very standard assumptions (independent and identically distributed data, finite variance, asymptotic data size) fastLSA expands some of the boundaries imposed by previous implementations. Written in optimized C, and utilizing POSIX threads, fastLSA is a software package can calculate pairwise LSA statistics and p-value bounds for ten thousand time series with time-steps measuring in the thousands.
- Linux / WIndows /MacOsX
:: MORE INFORMATION
BMC Genomics. 2013;14 Suppl 1:S3. doi: 10.1186/1471-2164-14-S1-S3. Epub 2013 Jan 21.
Expanding the boundaries of local similarity analysis.
Durno WE, Hanson NW, Konwar KM, Hallam SJ.