On 18/07/11 13:04, David Eccles (gringer) wrote: > It's also slower than sebhtml/ray (Total: 25 seconds vs 14 seconds; 57 > seconds if I include asserts and debug symbols). I presume this is > because of the additional complexity for getOutgoingEdges / getLastCode > (it needs to re-calculate the last base each time by iterating through > the k-mer). Perhaps the k-mer should store the last base as well as the > first....
So, I implemented a different k-mer format that included a last-base (reducing the max k-mer size for the same number of bytes by 1), and have worked out that this speed difference isn't only because of the re-calculation: --- With colour-space matching --- Sequence partitioning: 1 seconds K-mer counting: 4 seconds Coverage distribution analysis: 0 seconds Graph construction: 8 seconds Edge purge: 0 seconds Selection of optimal read markers: 6 seconds Detection of assembly seeds: 0 seconds Estimation of outer distances for paired reads: 1 seconds Bidirectional extension of seeds: 0 seconds Merging of redundant contigs: 1 seconds Generation of contigs: 0 seconds Scaffolding of contigs: 0 seconds Total: 21 seconds --- With base-space matching --- Sequence partitioning: 0 seconds K-mer counting: 3 seconds Coverage distribution analysis: 0 seconds Graph construction: 5 seconds Edge purge: 0 seconds Selection of optimal read markers: 3 seconds Detection of assembly seeds: 1 seconds Estimation of outer distances for paired reads: 0 seconds Bidirectional extension of seeds: 1 seconds Merging of redundant contigs: 0 seconds Generation of contigs: 0 seconds Scaffolding of contigs: 0 seconds Total: 13 seconds -- David ------------------------------------------------------------------------------ AppSumo Presents a FREE Video for the SourceForge Community by Eric Ries, the creator of the Lean Startup Methodology on "Lean Startup Secrets Revealed." This video shows you how to validate your ideas, optimize your ideas and identify your business strategy. http://p.sf.net/sfu/appsumosfdev2dev _______________________________________________ Denovoassembler-users mailing list Denovoassembler-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/denovoassembler-users