Singleton sequences are identified by sequence identifiers for SRA003728, with the size (in bp) and quality (% identity) of the alignment between the 454 sequence and the Sanger sequence.
The optional and last step in this example groups and counts all sequence identifiers, returning a list of all identified pairs located nearby and their count.
Notably, among the 9673 HPRD identifiers involved in PPIs, 119 were identified to correspond to obsolete nucleotide sequence identifiers, 4 corresponded to non protein-coding RNAs, while 16 were replaced by new nucleotide sequence identifiers; due to this updating, in three cases, two HPRD identifiers were assigned to the same nucleotide sequence identifier.
The noticeable differences between patent and journal extractions in Table 2 are that the later produce fewer compounds but more sequence identifiers.
The reason patents produce less human sequence identifiers is that smaller target classes, outside the "big-nine" listed in Methods, are not prioritised for extraction.
GVKBIO databases are populated with explicit relationships between compounds, assays and sequence identifiers that have been manually extracted from journals and patents on a large scale [3].
This currently includes out-links to PubMed for the scientific literature, MicroPatent [44] for patent documents, Entrez Gene for sequence identifiers, and PubChem CIDs as well as ChemSpider IDs for compound matches.
The reasons for less total sequence identifiers in patents are not only the target class restriction but also because activity data against non-human proteins is more commonly included in journal articles than patents.
WOrld of Molecular BioAcTivity (WOMBAT) is a commercial database that includes specific links between compounds and sequence identifiers for the proteins against which these compounds have been shown to be active [19].
These keys pointed to a list of template sequence identifiers.
(sequence identifiers can be found in Table S2).
