Biotech Values

[RockRat]

This patent is a pretty good overview of this totality -- though, as you say, part of it is in trade secrets -- if you can understand it.

WRT algorithms, they throw the whole associational rulebook at the problem: "In one embodiment the pattern is identified by linear discriminant, nearest neighbor, statistical classifier, neutral(sic) net, decision tree, decision rules or association rules analysis." Wonder if they really use all of these methods or are just throwing them out there to confuse the competition.

Head down to 0324 for "Computational Tools to Characterize Glycoprotein Mixtures"

Some off the shelf glycan degrading enzymes are mentioned, as are agents for chemical cleavage. All with their strengths and weaknesses, such that combinations have to be used . . .

Other fun stuff: different makes of reactors come up with different results despite use of the same control parameters, due apparently to the ways they agitate! Luckily, this seems only to affect yield.

http://appft.uspto.gov/netacgi/nph-Parser?Sect1=PTO1&Sect2=HITOFF&d=PG01&p=1&u=%2Fnetahtml%2FPTO%2Fsrchnum.html&r=1&f=G&l=50&s1=%2220100144553%22.PGNR.&OS=DN/20100144553&RS=DN/20100144553

Jeez, you're the guy who never sleeps, what am I doing?

Regards, RockRat