Chemistry Reference
In-Depth Information
Computer-intensive calculations can be solved using cloud, high performance and
grid approaches. Effectively, hybrid technologies should furnish advantages of
high performance, cloud and grid computing. Virtual library
construction/screening, molecular dynamics simulations and quantum mechanical
calculations can be performed using the afore-mentioned approaches.
The usage of cloud computing in molecular modeling for large-scale and data-
intensive applications appears to be compelling (generation of combinatorial
databases, analysis of huge genome sequencing/genomics projects, virtual
screening of millions of compounds).
However, in order to perform cloud computing, many technical/tedious tasks
needs performing (installation, launching, monitoring, and terminations). This
expertise is often lacking and may be one of the reasons for the infancy of
molecular modeling applications. There are not currently a corresponding large
number of publications using the CC protocol. Front-ends still needs to be
developed to facilitate user access.
Due to the emergence of virtualization and open source web protocols, encryption
to increase confidence cloud computing is a growing recent phenomenon.
However, data security and cost remains an issue for potential users. Cloud
computing appears to be past the peak of inflated expectations for emerging
technologies. However, nothing is perfect and backup plans need to me made as
well as robust architected solutions in order to avoid vendor lock-in.
POST-PROCESSING
Some of the new protocols use poses generated from techniques including
docking, scoring and virtual screening. Different procedures can be used to
improve (post-processing) the accuracy of the generated poses. Some of these
techniques include the following [42, 342, 356].
1)
Structural filtration from experimental binding.
2)
Interactions and visual inspection to repair deficiencies of scoring
functions.
