The NCBI non-redundant protein database nr was searched using the normal one-direction blastp with a cutoff E-value 1E-4.
The TAIR9 proteome database was searched using Mascot v2.2, allowing one missed tryptic cleavage, +2 and +3 charge states, a peptide tolerance of 20 ppm, and MS/MS ion tolerance of 0.6 Da.
In the network model for each species, regulatory motifs were searched using the ortholog groups, and the overlapped ones among the species were identified as evolutionarily conserved ones.
To undertake an evolutionary analysis of Su var)3-9 SET genes in land plants, three other completely sequencing plant genomes and one algal genome were searched using multiple representatives of Su var)3-9 SET proteins in Arabidopsis thaliana as queries.
More importantly, it provides a score which would actually be achieved during a search using one of the two models, while the other would score even higher.
Gene cluster coordinates were also searched using the reverse orientation for one of the two genomes.
In addition, the PubMed and SciFinder databases were also searched using the same search terms, resulting in one additional article.
Twenty-one databases were electronically searched using database-specific search strings incorporating "yoga", "musculoskeletal", "back pain", "arthritis" and "random".
Since the non-indexed search uses 10 20 nodes while the indexed search uses one node, non-indexed search may use a lot more total CPU time than indexed search even when the non-indexed wall-clock time is less.
Four electronic databases were searched by one reviewer (CG) using a systematic, comprehensive and reproducible search strategy.
