Text mining

Explosion

exponential increase

some things are constant

“graph calculus”

=

~45 seconds per paper

Information retrieval

find the relevant papers

user-specified query

“yeast AND cell cycle”

stemming

dynamic query expansion

ranking

Mitotic cyclin (Clb2)-bound Cdc28 (Cdk1 homolog) directly phosphorylated Swe1 and this modification served as a priming step to promote subsequent Cdc5-dependent Swe1

hyperphosphorylation and degradation

no tool will find that

Entity recognition

identify the substance(s)

Mitotic cyclin (Clb2)-bound Cdc28 (Cdk1 homolog) directly phosphorylated Swe1 and this modification served as a priming step to promote subsequent Cdc5-dependent Swe1

hyperphosphorylation and degradation

Mitotic cyclin (Clb2)-bound Cdc28 (Cdk1 homolog) directly phosphorylated Swe1 and this modification served as a priming step to promote subsequent Cdc5-dependent Swe1

hyperphosphorylation and degradation

Pafilis, O’Donoghue, Jensen et al., Nature Biotechnology, 2009

comprehensive lexicon

orthographic variation

“black list”

manual correction

still too much to read

Information extraction

formalize the facts

co-occurrence

global statistical analysis

NLPNatural Language Processing

parsing individual sentences

Gene and protein names

Cue words for entity recognition

Verbs for relation extraction

[nxexpr The expression of [nxgene the cytochrome genes [nxpg CYC1 and CYC7]]]is controlled by[nxpg HAP1]

Mitotic cyclin (Clb2)-bound Cdc28 (Cdk1 homolog) directly phosphorylated Swe1 and this modification served as a priming step to promote subsequent Cdc5-dependent Swe1

hyperphosphorylation and degradation

store in a database

then the fun begins :-)

Acknowledgments

NLP pipeline– Jasmin Saric– Rossitza Ouzounova– Isabel Rojas– Peer Bork

Reflect– Heiko Horn– Sune Frankild– Evangelos Pafilis– Sven Haag– Michael Kuhn– Peer Bork– Reinhardt Schneider– Sean O’Donoghue