Download presentation
Presentation is loading. Please wait.
1
Exploring Chemical Space with Computers—Challenges and Opportunities Pierre Baldi UCI
2
Chemical Informatics Historical perspective: physics, chemistry and biology Understanding chemical space Small molecules (systems biology, chemical synthesis, drug design, nanotechnology)
3
Chemical Space StarsSmall Mol. Existing10 22 10 7 Virtual010 60 (?) Access Difficult“Easy” Mode IndividualCombinatorial
4
Chemical Space
5
Chemical Informatics Historical perspective: physics, chemistry and biology Understanding chemical space Small molecules (systems biology, chemical synthesis, drug design, nanotechnology) Predict physical, chemical, biological properties (classification/regression) Build filters/tools to efficiently navigate chemical space to discover new drugs, new galaxies, etc.
6
Methods Spetrum: Schrodinger Equation Molecular Dynamics Machine Learning (e.g. SS prediction)
7
Chemical Informatics Informatics must be able to deal with variable-size structured data Graphical Models (Recursive) Neural Networks ILP GA SGs Kernels
8
Two Essential Ingredients 1. Data 2. Similarity Measures Bioinformatics analogy and differences: Data (GenBank, Swissprot, PDB) Similarity (BLAST)
9
Data Mutag (Mutagenicity) 200 compounds (125/63), mutagenicity in Salmonella PTC (Predictive Toxicity Challenge) A few hundred compounds, carcinogenicity (FM,MM,FR,MR) NCI (Anti-cancer activity) 70,000 compounds screened for ability to inhibit growth in 60 human tumor cell lines Alkanes (Boiling points) All 150 non-cyclic alkanes (C n H 2n+2 ) with n<11 and their boiling points ([- 164,174]) Benzodiazepines (QSAR) 79 1,4-benzodiazepines-2-one, affinity towards GABA A ChemDB 7M compounds
10
Similarity Rapid Searches of Large Databases Predictive Methods (Kernel Methods) Why it is not hopeless?
11
Similarity Rapid Search of Large Databases Protein Receptor (Docking) Small Molecule/Ligand (Similarity) Small Molecule/Ligand (Similarity) Predictive Methods (Kernel Methods) Why it is not hopeless OrganicChemicals
12
Linear Classifiers
13
Classification Learning to Classify Limited number of training examples (molecules, patients, sequences, etc.) Learning algorithm (how to build the classifier?) Generalization: should correctly classify test data. Formalization X is the input space Y (e.g. toxic/non toxic, or {1,- 1}) is the target class f: X → Y is the classifier.
14
Classification Fundamental Point: f is entirely determined by the dot products x i,x j measuring the similarity between pairs of data points
15
Non Linear Classification (Kernel Methods) We can transform a nonlinear problem into a linear one using a kernel.
16
Non Linear Classification (Kernel Methods) We can transform a nonlinear problem into a linear one using a kernel K. Fundamental property: the linear decision surface depends on K(x i,x j )= (x i ), (x j ) . All we need is the Gram similarity matrix K. K defines the local metric of the embedding space.
17
Similarity: Data Representations NC(O)C(=O)O
18
Molecular Representations 1D: SMILES strings 2D: Graph of bonds 2.5D: Surfaces 3D: Atomic coordinates 4D: Temporal evolution
19
15 Total: 1D SMILES Kernel CCCCCCc1ccc(cc1O)O CCCCCc1ccc(cc1)CO
20
2D Molecule Graph Kernel For chemical compounds atom/node labels: A = {C,N,O,H, … } bond/edge labels: B = {s, d, t, ar, … } Count labeled paths Fingerprints (CsNsCdO)
21
Similarity Measures
22
3D Coordinate Kernel 1.4 A 2.0 A 2.8 A 3.4 A 4.2 A
23
Example of Results
24
Results
27
Example of Results
28
Summary Derived a variety of kernels for small molecules State-of-the-art performance on several benchmark datasets 2D kernels slightly better than 1D and 3D kernels Many possible extensions: 2.5D kernels, isomers, etc… Need for larger data sets and new models of cooperation in the chemistry community Many open (ML) questions (e.g. clustering and visualizing 10 7 compounds, intelligent recognition of useful molecules, information retrieval from literature, docking, prediction of reaction rates, matching table of all proteins against all known compounds, origin of life) Chemistry version of the Turing test
29
ChemDB 7M compounds (3.5M unique) Commercially available PostgreSQL/Oracle Annotation (Experimental, Computational) Searchable Web interface Similarity, in silico reactions
30
Acknowledgements Informatics Liva Ralaivola J. Chen S. J. Swamidass Yimeng Dou Peter Phung Jocelyne Bruand Funding NIH NSF IGB Pharmacology Daniele Piomelli Chemistry G. Weiss J. S. Nowick R. Chamberlin
31
New Questions Predict drug-like molecules? toxicity? New Strategies How can we search efficiently? Intelligently? New data structures and algorithms Optimizing old structures How can we understand this much data? Cluster and visualize millions of data points Define commercially accessible space. Are there other useful things we can do with this? Discover new polymers, etc. Wonder about the origin of life. Combinatorially combine all known chemicals.
32
Acknowledgements Jocelyne Bruand Peter Phung Liva Ralaivola S. Joshua Swamidass Yimeng Dou NIH/NSF/IGB Questions
33
Docking Database of potential drugs 6 million small molecules … Query: Binding Site of Protein Scoring Function & Efficient Minimizer
34
Some Targets P53 (Luecke) ACCD5 (Tsai) IMPDH, PPAR, etc. (Luecke) HIV Integrase (Robinson)
35
P53
36
Drug Rescue of P53 Mutants
37
Docking → ChemDB ~6 million commercially available compounds Searchable, annotated, downloadable. Other Databases: Cambridge Structural Database ChemBank PubChem
38
Chemical Toxicity Prediction By Kernel Methods Jonathan Chen S Joshua Swamidass The Baldi Lab
39
Data Flow Toxicity State List Predictions Gram Matrix 4Yes2No3Yes 1No IDToxic? Kernel Linear Classifier
40
Results
41
Example of Results Kernel/Method Mutag MM FM MR FR Kashima (2003) 89.1 61.0 61.0 62.8 66.7 Kashima (2003) 85.1 64.3 63.4 58.4 66.1 1D SMILES spec. 84.0 66.1 61.3 57.3 66.1 1D SMILES spec+ 85.6 66.4 63.057.6 67.0 2D Tanimoto 87.8 66.4 64.2 63.7 66.7 2D MinMax 86.2 64.0 64.5 64.5 66.4 2D Tanimoto, l = 1024, b = 1 87.2 66.1 62.4 65.7 66.9 2D Hybrid l = 1024, b = 1 87.2 65.2 61.9 64.2 65.8 2D Tanimoto, l = 512, b = 1 84.6 66.4 59.9 59.9 66.1 2D Hybrid l = 512, b = 1 86.7 65.2 61.0 60.7 64.7 2D Tanimoto, l = 1024 + MI 84.6 63.1 63.0 61.9 66.7 2D Hybrid l = 1024 + MI 84.6 62.8 63.7 61.9 65.5 2D Tanimoto, l = 512 + MI 85.6 60.1 61.0 61.3 62.4 2D Hybrid l = 512 + MI 86.2 63.7 62.7 62.2 64.4 3D Histogram 81.9 59.8 61.0 60.8 64.4
42
Chemical Informatics Historical perspective: physics, chemistry and biology Understanding chemical space Small molecules (systems biology, chemical synthesis, drug design, nanotechnology) Catalog Predict physical, chemical, biological properties Build filters/tools to efficiently navigate chemical space to discover new drugs, new galaxies, etc.
43
Datasets
44
Small Molecules as Undirected Labeled Graphs of Bonds atom/node labels: A = {C,N,O,H, … } bond/edge labels: B = {s, d, t, ar, … }
45
Chemical Informatics Historical perspective: physics, chemistry and biology Understanding chemical space Small molecules (systems biology, chemical synthesis, drug design, nanotechnology) Bioinformatics analogy: Catalog (GenBank) Search (BLAST) Predict physical, chemical, biological properties Build filters/tools to efficiently navigate chemical space to discover new drugs, new galaxies, etc.
46
Chemical Informatics Historical perspective: physics, chemistry and biology Understanding chemical space Small molecules (systems biology, chemical synthesis, drug design, nanotechnology) Bioinformatics analogy: Catalog (GenBank) Search (BLAST) Predict physical, chemical, biological properties Build filters/tools to efficiently navigate chemical space to discover new drugs, new galaxies, etc.
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.