This can't all be good stuff: a paper I'm reading noted that "Just a few keywords (linkage, mapping,
SNP, genomewide association) identified 6866 articles in the PubMed database published in 2007 alone."

(Before you get too depressed, note that this means the total of each single keyword search, not a search for all 4 terms at once.)

Just for kicks, I tried my field. On Pubmed, there are 3664 papers published in 2009 alone that come up when I search for the term "cell cycle" (using quotes so that the search is for both words used together). If I limit my search to "cell cycle" and yeast, there are 242 papers from 2009. If I search for "transcription factors" and "yeast", there are 174 papers from 2009. "Cell cycle" and "transcription" brings up 699 papers.

A search for "systems biology" brings up 83 2009 review papers alone, and 744 total papers in 2009.

Trying other fields: a search for "genome-wide association study" turns up 388 hits from 2009. Searching for "human", "genome", "selection" brings up 308 papers from 2009.

The lesson here is that there is no time to waste reading lousy or irrelevant papers; just filtering things out properly takes a big chunk of time. You just have to be ruthlessly selective.