About Us

Committed to speeding innovation for science-driven organizations, from product ideation through commercialization.

Events

Visit us at the Fifth international workshop on Combinatorial Materials Research

June 11 - 12, 2013

Come and hear us speak at theFifth international workshop on Combinatorial Materials Research held at the Flanders Materials Centre in Ghent, Holland.

We are pleased to announce that David Nicolaides , Principal Field Applications Scientist, Accelrys will be presenting a poster entitled "Am I making a difference?" on 11th June between 1.45 - 3.00pm.

Details of the presentation can be found below:

Poster Presentation: "Am I making a Difference?"
Presentation details: Tuesday 11th June between 1.45 - 3.00pm.
Presenter: David Nicolaides , Principal Field Applications Scientist, Accelrys

Abstract
Developing a new material is not often easy, and the pressures of time-to-market can make it even harder.  We discuss two examples in the data mining part of the HTE workflow, where simple data mining methods reach false conclusions which have luckily been overturned when the data is re-mined with more thought, time, and expense.
 The first example involves high-throughput cell engineering experiments.  Here a naïve regression analysis seem to be telling the scientist that increasing the oxygen flow rate will increase the total amount of cells produced in her bioreactor array.  A more thorough regression analysis in fact reveals the truth; that some other unknown factor is at work.
The second example involves creating catalysts for hydrodesulfurization (HDS), where HTE methods are able to explore the high-dimensional parameter spaces more quickly and thoroughly than sequential methods.  However, these parameter variations don’t necessarily represent the only impact on HDS performance; variations in the latter can be mistakenly interpreted as being due to the former.  Here the naïve analyses are familiar (because they are easy to use and interpret) methods such as ANOVA and box plots.  The more complex analysis is factor analysis, which we describe in detail.
Both of these examples share the broad issue that the scientist assumes that things behave differently because they have made them behave differently.  This assumption is simply not true, and it must be tested in the HTE arena aggressively.

Conference Website >>>

We look forward to seeing you in Ghent!

Browse By: