TY - JOUR
T1 - Predictive models for protein crystallization
AU - Rupp, Bernhard
AU - Wang, Junwen
N1 - Funding Information:
We thank the current and past members of the TB Structural Genomics Consortium crystallization facility team (B.W. Segelke, H.I. Krupka, B.S. Schick, T. Lekin, J. Schafer, and D. Toppani) for populating the crystallization database. K.A. Kantardjieff, CSUF, has provided assistance with statistical data analysis and manuscript revisions. The cloning and protein production facilities under J. Perry, C. Goulding, and D. Eisenberg (UCLA); J.C. Sacchettini (Texas A&M University); T. Terwilliger, M. Park, C.-Y. Chang, and G. Waldo (LANL) have supplied a steady flow of proteins used in the crystallization experiments. Li Chen (RCSB Rutgers) has helped in extracting information from the PSI target database. LLNL is operated by University of California for the US DOE under contract W-7405-ENG-48. This work was funded by NIH P50 GM62410 (TB Structural Genomics) centre grant and produced with support of the Reiss Bar, Vienna, Austria.
PY - 2004/11
Y1 - 2004/11
N2 - Crystallization of proteins is a nontrivial task, and despite the substantial efforts in robotic automation, crystallization screening is still largely based on trial-and-error sampling of a limited subset of suitable reagents and experimental parameters. Funding of high throughput crystallography pilot projects through the NIH Protein Structure Initiative provides the opportunity to collect crystallization data in a comprehensive and statistically valid form. Data mining and machine learning algorithms thus have the potential to deliver predictive models for protein crystallization. However, the underlying complex physical reality of crystallization, combined with a generally ill-defined and sparsely populated sampling space, and inconsistent scoring and annotation make the development of predictive models non-trivial. We discuss the conceptual problems, and review strengths and limitations of current approaches towards crystallization prediction, emphasizing the importance of comprehensive and valid sampling protocols. In view of limited overlap in techniques and sampling parameters between the publicly funded high throughput crystallography initiatives, exchange of information and standardization should be encouraged, aiming to effectively integrate data mining and machine learning efforts into a comprehensive predictive framework for protein crystallization. Similar experimental design and knowledge discovery strategies should be applied to valid analysis and prediction of protein expression, solubilization, and purification, as well as crystal handling and cryo-protection.
AB - Crystallization of proteins is a nontrivial task, and despite the substantial efforts in robotic automation, crystallization screening is still largely based on trial-and-error sampling of a limited subset of suitable reagents and experimental parameters. Funding of high throughput crystallography pilot projects through the NIH Protein Structure Initiative provides the opportunity to collect crystallization data in a comprehensive and statistically valid form. Data mining and machine learning algorithms thus have the potential to deliver predictive models for protein crystallization. However, the underlying complex physical reality of crystallization, combined with a generally ill-defined and sparsely populated sampling space, and inconsistent scoring and annotation make the development of predictive models non-trivial. We discuss the conceptual problems, and review strengths and limitations of current approaches towards crystallization prediction, emphasizing the importance of comprehensive and valid sampling protocols. In view of limited overlap in techniques and sampling parameters between the publicly funded high throughput crystallography initiatives, exchange of information and standardization should be encouraged, aiming to effectively integrate data mining and machine learning efforts into a comprehensive predictive framework for protein crystallization. Similar experimental design and knowledge discovery strategies should be applied to valid analysis and prediction of protein expression, solubilization, and purification, as well as crystal handling and cryo-protection.
KW - High throughput crystallization
KW - Machine learning
KW - Predictive models
KW - Statistical analysis
KW - Structural genomics
UR - http://www.scopus.com/inward/record.url?scp=4344704198&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=4344704198&partnerID=8YFLogxK
U2 - 10.1016/j.ymeth.2004.03.031
DO - 10.1016/j.ymeth.2004.03.031
M3 - Article
C2 - 15325656
AN - SCOPUS:4344704198
SN - 1046-2023
VL - 34
SP - 390
EP - 407
JO - Methods
JF - Methods
IS - 3
ER -