Chemistry Reference
In-Depth Information
alternatives, hybrid MD/normal mode analysis, constrained geometrical
simulations, and random walk rotable bonds. The multiple protein structure
method (MPS) incorporates flexibility to identify hot spots using MD.
Simulations of MD also used to study intramolecular conformational changes and
water mediated hydrogen bonds [749-787].
Many machine-learning approaches have been also used in the last decade for
protein interaction site prediction (Bayesian Network, Support Vector Machine,
Artificial Neural Networks (classification methods) and Ramdom Forests,
Conditional Ramdon Fields (sequential labelling methods). With the exponential
growth of protein sequence data, an exploration of new methods that predict
protein interaction sites based only on sequence information is becoming
increasingly important. MetaPis is a recent sequence-based Meta-server for
protein interaction site prediction [750].
Despite diverse difficulties involved, several dozen published inhibitors (small
molecules) have been reported to disrupt protein-protein interactions [787]. This
is quite promising.
Stem Cells
Stem cells are very special. They are able to differentiate and self-renew to
somatic cells (mature)
in vitro
as well as
in vivo.
They differ in culture, longevity
and variety of types they can generate. Induced or embryonic, the pluripotent stem
cells are most potent. Adult cells are restricted in differentiation potential.
Stem cells' ability to generate relevant cells in limitless supply makes them
biopharmaceutical attractive. Over forty years old, bone marrow transplantation is
the most established. Other stem cell treatments have progressed to the clinical
stage [413-417, 805].
The high cost and not well-understood regulatory pathways hampers the
development of some therapies. There are also risks of transplantation of
undifferentiated/malignant transformations. Alternatively, stems cells can be used
in conventional small molecule drug discovery. Here we find well-established
