Introduction

Find out what I'm up to in my post-academic career at www.gbarr.info

Until March 2010, I was a post-doctoral research fellow in the Theoretical Crystallography group, part of WestCHEM at the Department of Chemistry, University of Glasgow. The group was led by Professor Chris Gilmore. I have since left for pastures new, but maintain close links to the Department

MooAlong with Andy Parkin, I am jointly responsible for Chem-moo-stry - the Chemistry Cow! A unique member of the 2006 Edinburgh Cow Parade, Chem-moo-stry.org is a scientific bovine, interested in explaining science through art – displaying some of the modern, cutting-edge chem-moo-stry research from the cow-laborations between the Scottish Universities forming ScotCHEM.
After the public Cow Parade ended, Cow #64 was brought back to Glasgow University's Chemistry Department, restored and on display after being unveiled by Prof. Anne Glover, the Chief Scientific Advisor for Scotland. She has now moo-ved on to be displayed at the Glasgow Science Centre.

I was also involved in helping with IT support in the Chemistry Department, in particular working on special projects, including the new Chemistry Plasma Information Screen System, and other database and web-related systems.

Research Interests

I developed software that applies numerous types of cluster analysis and other multivariate statistical methods to problems in crystallography and chemistry, and use computer graphics to display the results in an intuitive and easy-to-interpret manner.

PolySNAP in action

PolySNAP is a software package designed to match and analyse powder diffraction patterns utilising their full profiles. The use of the full-profile allows for more flexible and accurate identification of samples, even when data quality is low or preferred orientation effects are significant.

The software provides an easy to use interface to several powerful and novel statistical methods to rank patterns in order of their similarity to any selected sample, allowing unknowns to be quickly identified. In quantitative mode, given a mixture pattern and potential pure phase patterns, it can identify which patterns are in the mixture, and quantify their proportions quickly and easily using a non-Rietveld based approach.

PolySNAP

PolySNAP
PolySNAP automates the matching procedure for high-throughput analysis. It allows for datasets of up to 1,500 patterns to be analysed in a single run, and provides highly flexible graphical output to summarise and visualise the results. This highlights any unusual data, and means that time is not wasted looking at the very many patterns that behave exactly as expected.
It can work with or without the provision of reference patterns, and includes features such as an automated report writer and a time/date stamped logfile to assist with audit trail procedures.

PolySNAP is distributed commercially by Bruker-AXS, as is another simpler, more focussed version of the program called PolySNAP M - click here for a table that describes the different versions of the program.
 
dSNAP
dSNAP takes the clustering and visualisation methods originally developed to analyse powder diffraction patterns, and applies  them to structural geometries mined from the Cambridge Structural Database.
This allows the user to interpret the underlying chemistry behind those classifications in a manner that diminishes the need for tedious manual comparisons. 
dSNAP is available for free download.


Selected Recent Publications

Configurational and conformational classification of pyranose sugars
Collins, A., Parkin, A., Barr, G., Dong, W., Gilmore, C.J., and Wilson, C.C. Acta. Cryst. B64, 57-65

Using the Cambridge Structural Database to validate powder structures
Barr, G., Dong, W., Gilmore, C.J, Kern, A., Parkin, A., and Wilson, C.C. Z. Kristallogr. Suppl. 26 (2007) 209-214

A Quick Method for the Quantitative Analysis of Mixtures. 1. Powder X-Ray Diffraction
Dong, W., Gilmore, C., Barr, G., Dallman, C., Feeder, N., and Terry, S. J. Pharm. Sci., (2007), 10.1002/jps.21142

Using cluster analysis to study transition-metal geometries: four-coordinate complexes with two salicylaldiminato or related ligands
Parkin, A., Barr, G., Collins, A., Dong, W., Gilmore, C.J., Tasker, P. and Wilson, C. C. Acta Cryst. (2007). B63, 612-620


Comparing entire crystal structures: structural genetic fingerprinting
Parkin, A., Barr, G., Dong, W., Gilmore, C.J., Jayatilaka, D., McKinnon, J.J., Spackman, M.A, and Wilson, C.C. CrystEngComm, (2007) 8, 613-714

- Featured in the July 2007 issue of RSC Chemical Science
 

- Cover Article for August 2007 issue of CrystEngComm.

 

The application of cluster analysis to identify conformational preferences in enones and enimines from crystal structural data

Collins, A., Barr, G., Dong, W., Gilmore, C.J., Middlemass, D.S., Parkin, A. and Wilson, C. C. Acta Cryst, (2007). B63, 469-476

Identifying structural motifs in inter-molecular contacts using cluster analysis
Part 2. Interactions of carboxylic acids with secondary amides
Collins, A., Parkin, A., Barr, G., Dong, W., Gilmore, C.J. and Wilson, C. C. CrystEngComm, (2007). 9, 222 - 227


A full list of publications may be found here.

Background

Ph.D. in Chemistry in 2003 at the University of Glasgow. First degree was also from Glasgow - an M.Sci. in Chemistry in 1999.  Undergraduate Fourth Year project was with Dr. Paul Mallinson on charge density studies of weak hydrogen bonds.
My interests include books, architecture and history (especially cinemas and cinema architecture, Glasgow and the West End - see the personal links on the right), and other more normal stuff.