Search results

Results come from paper abstracts only. In a re-run on reviewed papers the extraction recovered roughly 72–83% of the results a person had found; it is weakest for heterologous protein or enzyme expression and promoter or signal-sequence work. A strain is recorded for only about a third of results.