A model example of P
erhaps the most famous image of chemical modelling is that of James Watson and Francis Crick
standing next to the model of the 3D structure of the double helix of DNA that they had built at the Cavendish Laboratory in Cambridge. Six decades later, molecular modelling takes place virtually rather than with physical components painstakingly fitted together like a jigsaw. Lalitha Subramanian,
senior director and fellow at software informatics company Accelrys, believes one advantage of molecular modelling and simulation software is that researchers can view a system from different angles. She explains that such packages have been used successfully in several different ways, for example to understand experimental observations, or as a virtual screening of millions of possibilities to supplement experimental data. ‘It is also
used to test out various “what if” scenarios virtually before performing the experiment,’ she says, ‘giving the researcher a productive path forward.’ These scenarios provide a way
of addressing key questions such as ‘why is something happening’, ‘how is this affecting’, ‘why are the changes causing a property to enhance or decrease’ and, as Subramanian points out, modelling software gives chemists access to a pool of knowledge enriched with data of a different type from experimental data. ‘This type of key insight is very difficult to obtain from experiments alone,’ she says. ‘Over the years our customers have seen the value of modelling and simulations and some have avoided months of intensive laboratory screening. For others, it has helped in focusing their future experiments along the most productive path. For all, it provides deep insights into their system and process that
Chemical modelling software is revolutionising the way in which chemists approach their work, as Beth Sharp discovers
chemistry Optibrium’s StarDrop software in action
no experiments can provide,’ Subramanian adds.
Beyond the lab Drug discovery is so complex that scientists are limited in what they can achieve through experiment alone. Not only is time in which to conclude research limited, but in most circumstances they will want to investigate far more possibilities than can actually be explored in any practical way. Matt Segall, CEO of Optibrium, explains: ‘Early on in a project, it doesn’t necessarily make sense to put all your efforts into one small area of chemistry and generate lots of similar molecules and data. You will want to spread your risk across different chemistries and to explore potential back- up series and, from a modelling perspective, understand the structure-activity relationship in your data.’ This approach allows
Finding correlated variables by using Qlucore’s Omics Explorer 2.1 36 SCIENTIFIC COMPUTING WORLD individuals to look at the
data objectively and select the molecules they want to take forward, adds Segall, by making and testing virtual compounds or progressing molecules discovered in preliminary experimental data to secondary or in vivo studies. But, with many options available on the market, how can a researcher decide which modelling software best suits their needs? Segall believes that researchers are hesitant to use ‘black boxes that are given a molecule and shoot out a number, leaving the user to question immediately why that prediction has been made and what can be done to make improvements.’ Mere presentation of data is not sufficient, and so developers like Optibrium aim to deliver tools that effectively ‘guide’ decisions in the face of the statistical uncertainty that comes with any predictive model.
www.scientific-computing.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44