George Papadatos - Knime Tutorial

Day 4: KNIME Practical George Papadatos, ChEMBL group, EMBL-EBI

Francis Atkinson, ChEMBL group, EMBL-EBI

Outline

2

Introduction to KNIME

Basic components Desktop, nodes, dialogs, workflows

Demo Compound selection for focused screening

Read chemical data

Calculate properties

Apply drug- and lead- likeness filters

Remove nasty compounds

Pick diverse molecules

Visualize results and plot properties

Exercises 1 & 2 (hands-on)

12/12/2013 Resources for Computational Drug Discovery

Are there KNIME users among us?

Resources for Computational Drug Discovery 12/12/2013 3

What is KNIME?

KNIME = Konstanz Information Miner

Developed at University of Konstanz in Germany

Desktop version available free of charge (Open Source)

Modular platform for building and executing workflows using predefined components, called nodes

Core functionality available for tasks such as standard data mining, analysis and manipulation

Extra features and functionality available in KNIME through extensions from various groups and vendors

Written in Java based on the Eclipse SDK platform

4 12/12/2013 Resources for Computational Drug Discovery

KNIME resources

Web pages (documentation) www.knime.org | tech.knime.org | tech.knime.org/installation-0

Downloads knime.org/download-desktop

Community forum tech.knime.org/forum

Books and white papers knime.org/node/33079

Myself

[email protected]


What can you do with KNIME?

6

Data manipulation and analysis File & database I/O, sorting, filtering, grouping, joining, pivoting

Data mining / machine learning R, WEKA, KNIME, interactive plotting

Chemoinformatics Conversions, similarity, clustering, (Q)SAR analysis, MMPs, reaction

enumeration

Scripting integration R, Perl, Python, Matlab, Octave, Groovy

Reporting

So much more Bioinformatics, HTS & image analysis, network & text mining

Marketing, bid data and business analytics


Community contribution nodes

http://tech.knime.org/community

Chemoinformatics ChEMBL and ChEBI (EBI) SureChEMBL nodes coming soon!

CDK (EBI), RDKit (Novartis), Indigo (GGA), ErlWood (Eli Lilly), Enalos (NovaMechanics)

Bioinformatics HCS (MPI), NGS (Konstanz), Image analysis

Text mining Palladian

Integration Python, Perl, R, Groovy, Matlab (MPI), PDB web services client (Vernalis)


http://tech.knime.org/communityhttp://tech.knime.org/community

Installation & updates

8

Download and unzip KNIME No further setup required

Additional nodes after first launch

knime.ini contains arguments & parameters for launch

New software (nodes) from update sites http://tech.knime.org/update/community-contributions/release

Workflows and data are stored in a workspace /Users/georgep/knime/workspace_mac_new

C:\knime_2.8.2\workspace

Customization in: FilePreferencesKNIME


KNIME Workbench

9

workflow editor

console outline

tabs

Node description

node repository

workflow projects

favorite nodes

public server


Auto-layout Execute Execute all nodes

Node = basic processing unit of KNIME workflow which performs a particular task

Title

Icon

Input port(s) on the left of icon

Output port(s) on the right of icon

Status display (traffic lights) Red (not ready) Amber (ready) Green (executed)

Blue bar during execution

(with percentage or flashing)

Sequence number Right-click menu To configure and execute the node, display the output views, edit the node, and display data for the ports

KNIME nodes: Overview


Configuration menus for selected nodes

KNIME nodes: Dialogs

11

Explicit column type


Double click to configure

An example completed workflow

12

Workflows can be imported and exported as .zip files

With or without the underlying data

File Import KNIME workflow

File Export KNIME workflow


Any questions so far?


Compound selection for focused screening

1. Read chemical data

2. Calculate phys/chem properties

3. Apply drug- and lead-likeness filters

4. Apply more filters (e.g. remove solubility liabilities)

5. Apply substructural filters (PAINS subset)

6. Pick diverse molecules


The objective


First steps - I

Locate the directory with todays material

Copy and paste it to your desktop

You can take it with you too

Open the presentation file

Import the FocusedScreeningSelection.zip to KNIME

Menu File Import workflow to KNIME


1

2

3

First steps - II

Open a new workflow

Right click on the workflow projects area


1

2

3

Part 1: Reading chemical data


SDF Reader


1

2

3

5

4

.\data\SMDC_cleaned_nodups.sdf

Inspect the structures


Right click on the node

Molecule to RDKit


Part 2: Property-based filtering


Descriptor Calculation


1 2

3

Java Snippet


1

2

3

.\code\Lipinski.txt

Numeric Row Splitter


Inspect the Lipinski fails



Java Snippet


.\code\Oprea.txt 1

2

3

Inspect the Oprea fails




Inspect the Solubility fails


Part 3: Substructure-based filtering


Molecule to Indigo



File reader .\data\PAINS_clean_half.sdf


Query Molecule to Indigo


Inspect the SMARTS rules


Chunk Loop Start


Substructure Matcher


Loop End


Inspect matched structures



Reference Row Filter

Part 4: Diversity picking and plotting


RDKit Fingerprint


Inspect the fingerprints



RDKit Diversity Picker


2D/3D Scatterplot


Inspect the plot



Exercise 1

Read an sd file with drug information from ChEMBL

Inspect the structures and their properties

Select only drugs that were released after 1990 (First Approval)

Select only drugs that target human (Homo sapiens)

How many drugs remain now?

Save the workflow

Tips Open a new workflow

Use the SDF Reader node

Use the Numeric Row Splitter node to filter on First Approval >= 1990

Use the Nominal Value Row filter node to filter on Organism = Homo sapiens


Exercise 2

Continue from your previous workflow

Calculate MW and logP of the drug compounds

Generate a scatter plot of MW and logP

Can you see any compounds with high MW and logP?

Tips Use the Molecule to RDKit node

Use the RDKit Descriptor Calculator node

Include the SlogP and ExactMW descriptors

Use the 2D/3D Scatterplot node


Any questions? Last chance!


Conclusions

Compound selection for focused screening

Typical scenario

KNIME

Open and free

Data analysis

Chemoinformatics toolkits Erl Wood, RDKit, Indigo, CDK, etc.

Lots of other functionality

More advanced KNIME on Friday around lunch time


Further reading

Open data and tools


1. Irwin, J. J.; Sterling, T.; Mysinger, M. M.; Bolstad, E. S.; Coleman, R. G., ZINC:

A free tool to discover chemistry for biology. Journal of Chemical Information

and Modeling 2012 ASAP.

2. Saubern, S.; Guha, R.; Baell, J. B., KNIME workflow to assess PAINS filters in

SMARTS format. Comparison of RDKit and Indigo cheminformatics libraries.

Molecular Informatics 2011, 30, (10), 847-850.

3. Barnes, M. R.; Harland, L.; Foord, S. M.; Hall, M. D.; Dix, I.; Thomas, S.;

Williams-Jones, B. I.; Brouwer, C. R., Lowering industry firewalls: pre-

competitive informatics initiatives in drug discovery. Nature Reviews Drug

Discovery 2009, 8, (9), 701-708.

4. Berthold, M. R.; Cebron, N.; Dill, F.; Gabriel, T. R.; Ktter, T.; Meinl, T.; Ohl, P.;

Sieb, C.; Thiel, K.; Wiswedel, B., KNIME: The Konstanz Information Miner. In

Data Analysis, Machine Learning and Applications, Preisach, C.; Burkhardt, H.;

Schmidt-Thieme, L.; Decker, R., Eds. Springer: Berlin, 2008; pp 319-326.

5. Tiwari, A.; Sekhar, A. K. T., Workflow based framework for life science

informatics. Computational Biology and Chemistry 2007, 31, (5-6), 305-319.

Further reading

High throughput screening

Lead- and drug-likeness


1. Bajorath, J., Integration of virtual and high-throughput screening. Nature

Reviews Drug Discovery 2002, 1, (11), 882-894.

2. Harper, G.; Pickett, S. D.; Green, D. V. S., Design of a compound

screening collection for use in High Throughput Screening. Combinatorial

Chemistry & High Throughput Screening 2004, 7, (1), 63-70.

1. Chuprina, A.; Lukin, O.; Demoiseaux, R.; Buzko, A.; Shivanyuk, A., Drug- and

lead-likeness, target class, and molecular diversity analysis of 7.9 million

commercially available organic compounds provided by 29 suppliers. Journal of

Chemical Information and Modeling 2010, 50, (4), 470-479.

2. Lipinski, C. A., Lead- and drug-like compounds: the rule-of-five revolution. Drug

Discovery Today: Technologies 2004, 1, (4), 337-341.

3. Oprea, T. I.; Davis, A. M.; Teague, S. J.; Leeson, P. D., Is there a difference

between leads and drugs? A historical perspective. Journal of Chemical

Information and Computer Sciences 2001, 41, (5), 1308-1315.

Further reading

Physicochemical properties and drug discovery

Structural alerts in HTS


1. Brstle, M.; Beck, B.; Schindler, T.; King, W.; Mitchell, T.; Clark, T., Descriptors,

physical properties, and drug-likeness. Journal of Medicinal Chemistry 2002, 45,

(16), 3345-3355.

2. Hill, A. P.; Young, R. J., Getting physical in drug discovery: A contemporary

perspective on solubility and hydrophobicity. Drug Discovery Today 2010, 15,

(15/16), 648-655.

3. Leeson, P. D.; Springthorpe, B., The influence of drug-like concepts on decision-

making in medicinal chemistry. Nature Reviews Drug Discovery 2007, 6, (11), 881-

890.

1. Baell, J. B.; Holloway, G. A., New substructure filters for removal of Pan Assay

Interference Compounds (PAINS) from screening libraries and for their exclusion in

bioassays. Journal of Medicinal Chemistry 2010, 53, (7), 2719-2740.

2. Rishton, G. M., Reactive compounds and in vitro false positives in HTS. Drug

Discovery Today 1997, 2, (9), 382-384.

Further reading

Similarity and diversity


1. Ashton, M.; Barnard, J.; Casset, F.; Charlton, M.; Downs, G.; Gorse, D.; Holliday,

J.; Lahana, R.; Willett, P., Identification of diverse database subsets using

property-based and fragment-based molecular descriptions. Quantitative

Structure-Activity Relationships 2002, 21, (6), 598-604.

2. Bender, A.; Glen, R. C., Molecular similarity: a key technique in molecular

informatics. Organic and Biomolecular Chemistry 2004, 2, 3204-3218.

3. Gorse, A.-D., Diversity in medicinal chemistry space. Current Topics in Medicinal

Chemistry 2006, 6, (1), 3-18.

4. Maldonado, A.; Doucet, J.; Petitjean, M.; Fan, B.-T., Molecular similarity and

diversity in chemoinformatics: From theory to applications. Molecular Diversity

2006, 10, (1), 39-79.

5. Rogers, D.; Hahn, M., Extended-connectivity fingerprints. Journal of Chemical

Information and Modeling 2010, 50, (5), 742-754.

6. Schuffenhauer, A.; Brown, N., Chemical diversity and biological activity. Drug

Discovery Today: Technologies 2006, 3, (4), 387-395.

7. Willett, P.; Barnard, J. M.; Downs, G. M., Chemical similarity searching. Journal

of Chemical Information and Computer Sciences 1998, 38, (6), 983-996.

Day 4: KNIME Practical George Papadatos, ChEMBL group, EMBL-EBI

Francis Atkinson, ChEMBL group, EMBL-EBI

George Papadatos - Knime Tutorial

Documents

unzip knime

downloads knime

knime users

white papers knime

knime basic components

nodes node

additional nodes

data manipulation