|
"Drug
design is a huge challenge from every perspective, but
the people doing modeling are real heroes,” says
Shi-Yi Liu, vice president at Schrödinger. “They are
trying to truly understand intermolecular interactions
at the atomic level. And if you can understand things at
that level, you can project them onto different systems,
rather than having to reinvent the wheel each time.”
Liu and
her colleagues have chosen some of the very toughest of
these problems to address protein structure prediction
through new tools for homology modeling and
docking. The key to this, according to Liu, is lots of
iterations.
Schrödinger’s
staff, more than half of whom are Ph.D.’s, works
closely with clients to try and improve the software.
“When we put forth a method, it won’t immediately
work for every case,” Liu says. By better
understanding cases where a method has failed, the Schrödinger
scientists can improve upon their algorithm.
One of
Schrödinger’s unique tools is its Induced Fit
solution, which combines its Prime and Glide programs
for predicting ligand-induced conformational changes in
receptors. “When a drug binds to a protein, the
protein’s structure will change,” Liu explains.
“Our induced fit approach is a fairly significant step
toward understanding receptor flexibility.” But it
took an immense amount of work. “In the original
paper, we studied more than 20 systems and they all
behaved differently,” she says.
Case
studies such as the company’s HERG homology model and
the work of Debananda Das and colleagues (see
“Modeling New HIV Therapies,” page 21) prove that
the field is definitely making progress, even though
there is always frustration. “We are trying to
simulate a multitude of real-life factors with only a
few variables that can be coded up mathematically,”
Liu explains. “Often, it’s an approximation of an
approximation.”
In a
field like this one, progress usually comes in small
doses, and there are always big questions still waiting.
“I don’t think we are quite all the way there
yet,” Liu says. “We haven’t solved all the
problems, but I’d like to think our products are
getting us closer.”
|
