SYMBIOmatics: Synergies in Medical Informatics and Bioinformatics – exploring current scientific literature for emerging topics

Table 5 New bigrams in the BI journal corpus in recent years.

New in 2004–2005	Df	New in 2002–2003	Df	New in 2000–2001	Df
protein background	16	false discovery	41	microarray datum	268
method conclusion	12	discovery rate	40	microarray experiment	183
annotation method	11	datum background	40	expression profile	168
dataset result	11	microarray study	36	microarray data	161
array cgh	10	text mining	35	gene ontology	135
protein localization	10	association study	28	support vector	133
organism database	10	r package	26	vector machine	130
ontology database	10	normalization method	25	protein interaction	99
biocreative task	9	multiple testing	23	nucleotide polymorphism	76
entity recognition	9	ontology term	22	cdna microarray	73
splicing event	8	go term	21	microarray technology	73
name recognition	8	gene list	20	microarray gene	65
lowess normalization	8	human protein	20	differential expression	59
anatomy ontology	7	biomedical text	19	open source	54
novo sequencing	7	complex disease	19	biological network	50
task 2	6	microarray result	18	microarray analysis	48
task 1a	6	homo sapiens	18	widely used	48
venn diagram	4	named entity	17	gene selection	46
database identifier	4	synonymous codon	16	interaction datum	37
		gene clustering	16	system biology	34
		mammalian genome	16	interacting protein	33
		bioinformatics analysis	15	alternative splicing	32
		haplotype block	14	oligonucleotide microarray	29
		go annotation	13	related gene	27
		two dataset	13	web application	27
		expression result	13	biological sample	26
		marker gene	12	expression value	23
		dimensionality reduction	12	primer design	22

The table shows bigrams from the BI journal corpus that were new during the period 2004–2005 (col. 1 and 2), the period 2003–2004 (col. 3 and 4) and the period 2000–2001 (col. 5 and 6). All bigrams were selected and ranked according to their document frequency value (ref. to text), which had to be above 3. During the time 2000–2001 a large number of bigrams referring to microarray experiments emerged. "task 1a" and "task 2" are exclusively linked to BioCreAtive. "false discovery" refers to false discovery rate (FDR) in DNA microarray analysis.

ISSN: 1471-2105