3. STSS_SOW APP screening.pdf

Statement of Work

NCATS Screening Data Processing and Data Analysis Platform

May 12, 2023

TABLE OF CONTENTS

Contents

1.0 INTRODUCTION/OVERVIEW

1.1 BACKGROUND

1.2 SCOPE

1. Drug Combination Matrix Assay (CMA) Designer & Manager

2. Centralized Screening Data Processing Platform and Interface

3. Single Agent Screening Data Explorer and Visualization

4. Matrix Screening Data Explorer and Visualization

5. HCS Data Acquisition and Catalog

6. HCS Data Explorer and Visualization

7. Integration with NCATS Systems

2.0 OBJECTIVES

Module A – Drug Combination Matrix Assay Designer & Manager

Module B - Centralized Screening Data Processing Platform and Interface

Module C – Single Agent Screening Data Explorer and Visualization

Module D - Matrix Combination Screening Data Explorer and Visualization

Module E – Creation of New Database and Migration of Historical data

Module F – Integration with NCATS systems

3.0 REQUIREMENTS

3.1 TASKS

3.2 DELIVERABLES

1.0 INTRODUCTION/OVERVIEW

1.1 BACKGROUND

With an increasing number of approved and investigational drugs, the discovery of novel drug combinations through clinical trial-and-error is not feasible. To address this inefficiency, the NCATS scientists uses various drug screening formats across a variety of assay types for early stage drug discovery and repurposing projects. Along with these screening platforms, NCATS has created a variety of applications for processing, visualizing and analyzing this data. These applications enables NCATS scientists to quickly narrow down a long list of potential drugs and find those with the most potential to help patients.

NCATS seeks an upgraded set of these applications to improve the process by which NCATS researchers can rapidly test the effect of different drugs on cellular, molecular or biochemical processes that are relevant to a disease of interest. This solution will support all NCATS scientists screens It will also support the DPI Informatics team, responsible for analyzing the screening data that scientists can use the results of these screens to pursue testing of promising drugs in animals and, ultimately, humans.

1.2 SCOPE

NCATS seeks support in overhauling its screening data processing and analysis applications to remain at the forefront of this field. The successful implementation will be state-of-the-art, qualitative, and high throughput to support the highly integrated NCATS robotics systems. Due to these requirements, NCATS requires the vendor to provide a custom software solution in the following scope areas.

1. Drug Combination Matrix Assay (CMA) Designer & Manager Currently NCATS Compound Management (CoMa) uses a disparate set of tools to create the scripts and various other files needed to transfer compounds into assay plates in a matrix screening format. We need a centralized Matrix Assay Designer software that will create the screening combinations, scripts and plate maps needed by CoMa. These platform needs to be integrated with existing infrastructure so files and data required to conduct the combination screens are generated automatically and available to various groups who need to process and analyze matrix screening data.

2. Centralized Screening Data Processing Platform and Interface Creation of a centralized platform which is capable of taking raw screening data of any type and going through the steps to take raw data to the final normalized data. The platform should allow for annotation of the normalized data with compound information to allow for curve fitting (single agent) or synergy calculation (matrix data). This platform should utilize existing code and functions present in the Java-based qHTS console client and NCATS ncgcmatrix R package.

3. Single Agent Screening Data Explorer and Visualization NCATS users doing qHTS Single Agent Screening need to visualize compound dose response curves and plate heatmaps from a specific experiment or do comparisons across multiple experiments. Creation of an interface that allows for data browsing, dose response curve imaging, plate heatmap imaging, experimental comparisons and exporting of images and data will support NCATS scientific efforts. This interface will replace the existing Java-based qHTS client.

4. Matrix Screening Data Explorer and Visualization NCATS users doing drug combination Screening need to visualize heatmaps of activity data and calculated synergy metrics for each combination. Creation of an interface that allows for viewing of matrix data, experimental annotation, plate heatmap images, and exporting of images and data will support NCATS scientific efforts. This interface will replace the existing matrix client website at matrix.ncats.nih.gov.

5. HCS Data Acquisition and Catalog Data acquisition and cataloging of is a broad area of work scope where NCATS envisions a unique User-interface that provides straightforward and intuitive data browsing. The software platform should integrate and streamline the disparate existing systems and manually intensive processes to transfer data to and from each platform.

6. HCS Data Explorer and Visualization NCATS users analyzing High Content Screening (HCS) often want to compare a compound or small subset of compounds across assays using the existing CMS tools.

However, other potential views of the data that could help researchers better understand and uncover novel insights from the data are by the assay, animal model, curve classification, or mode of action (MOA). This interface should integrate with scope items C and D to allow for viewing of HCS data and other assay data in the same interface.

7. Integration with NCATS Systems The above applications/software will integrate all existing software so the applications are interoperable with each other and will increase the efficiency of the existing screening program. Any areas where the CMA system can leverage existing NCATS systems, tools, or management functions (e.g. identities) should be used.

2.0 OBJECTIVES

The objective of year two is to create and deploy various applications which are currently in a development state and to create and deploy all the back-end infrastructure required for these new applications. Additionally, migration of all historical data from the existing

NCATS ProbeDB Oracle database will be required to ensure that NCATS scientist can view historical and future data side by side.

Module A – Drug Combination Matrix Assay Designer & Manager

1. Replace the functionality provided by existing legacy solutions outlined in Table 1:

Existing inventory of matrix combination screening software tools.

Table 1: Existing inventory of matrix combination screening software tools

Application Function Language CoMa Admin Tools Website

This web application can retrieve compounds from the compound store https://nimoy2016.ncats.nih.gov/coma/compoundstore/placestoreo rder.aspx

ASP.NET

C#

Matrix Source Plate Generator

This desktop application is used to generate the source plates, script files, and data map files to be used by the robotic platform.

WPF

C#

Barcode Script This Labview application adds the appropriate barcode to each individual map file before it can be uploaded to the data analysis tool

LabView

Matrix Data Analysis Tools

This desktop application requires the script and map files to be uploaded manually before it can process the matrix screening data

Java, R

Matrix Screening Client

This web application allows researchers to view the results of their screens after their data has been analyzed and is publicly available.

https://matrix.ncats.nih.gov/

Java

2. Designing and documenting new architecture for the integrated Matrix Combination Screening Platform

3. Storing the data maps of the destination plates in a central location that is accessible by other applications

4. Developing application components responsible for creating the matrix source plates required for combination screening

5. Developing application components responsible for creating the scripts that transfer compounds from the source plates to the destination plates

6. Developing application components that create the map files that correspond to the contents of the new destination plates

Module B - Centralized Screening Data Processing Platform and Interface

1. Development of a tool which allows for users to parse raw data from various plate reader output files

1.1. Viewlux

1.2. Pherastar

1.3. Acumen

1.4. Phenix

1.5. Envision

https://nimoy2016.ncats.nih.gov/coma/compoundstore/placestoreorder.aspx https://nimoy2016.ncats.nih.gov/coma/compoundstore/placestoreorder.aspx https://matrix.ncats.nih.gov/

1.6. Readers not yet purchased by NCATS

2. Development of a tool which can take raw data and execute functions needed for downstream analysis

2.1. Transform raw data into activity data

2.2. Transform activity data into normalized data

2.3. Apply plate level corrections to normalized data

2.4. Calculate plate level QC statistics on data

2.5. Exclude plates or exclude wells based on QC data

3. Development of tool which can take final normalized data and annotate with compound data. This data is currently stored in the Oracle ProbeDB database.

4. Development of a tool which can take normalized and compound annotated data and apply calculations specific to single agent drug screening

4.1. Pivot and calculate 4-parameter curve fit and metrics including

4.1.1. IC50 (concentration at which 50% inhibition of the response is achieved)

4.1.2. AC50/EC50 (concentration at which 50% of maximum response is achieved)

4.1.3. Area under the Curve (AUC)

4.1.4. Hill slope

4.1.5. Maximum response

4.1.6. NCATS Curve Response Class

4.1.7. Other curve based metrics

5. Development of an application which can connect each of the outputs of points 1-4 and report all settings and selections made as well as append assay data

6. Development of a tool which can take normalized and compound annotated data and apply calculations specific for drug combination data analysis including

6.1. Highest Single Agent (HSA) model drug synergy

6.2. Bliss model drug synergy

6.3. Loewe model drug synergy

6.4. Lou Staudt 3x3 model drug synergy

6.5. Novel drug synergy methods not yet existing

Module C – Single Agent Screening Data Explorer and Visualization

1. Development of visualization tool that allows users to view and export in a single interface

1.1. Dose response curves

1.2. Protocol information

1.3. Compound information

1.4. Screening results as heatmaps

1.5. Compound comparisons by user defined filters

1.6. Compound comparisons with other datasets outside of the matrix screening data

2. Scalable interactive visualization, filtering, and overlaying of current matrix screening data

2.1. A UI that can scale to 1B plus compounds/concentration interactions

2.2. All interactions labeled, selectable, and filterable in all interactions

2.3. Ability to place multiple different compound matrices side by side or overlay them to compare responses.

2.4. A UI that enables multiple assay outputs to be dynamically switched between

3. Simple compare and overlay of drugs/concentrations of interest

3.1. When combinations of compounds/concentrations are identified, allow them to be selected and then a new "subset" visualization can be built based only on promising targets selected by the user

4. Dynamic Switching of data organization as sorted by:

4.1. Assay,

4.2. Animal model

4.3. MOA information

4.4. Or any other user defined metadata

5. Interactive overlays of:

5.1. Multiple outcome measures

5.2. Curve classification

5.3. Matrix response

5.4. Delta Bliss

5.5. Dose response curves

6. Inside the data exploration tool, enable discovery of foundational information for the screen

6.1. Protocol information

6.2. Animal model

6.3. Cell line

6.4. Compound information

6.5. Outcome measure descriptions

Module D - Matrix Combination Screening Data Explorer and Visualization

1. Development of website which allows for the following functions of the following data:

1.1. Data Viewing and Export

1.1.1. drug synergy calculations

1.1.2. drug synergy heatmaps

1.1.3. compound structural images and mechanism annotations

1.1.4. plate heatmaps

1.1.5. experimental conditions/details

1.1.6. all data above should be able to be exported

1.2. Data Structure

1.2.1. project folders should be able to be created with annotation of project level information

1.2.2. experiments should be able to be associated with a project folder

1.2.3. experiment annotation should be possible to note cell line, conditions, assay used, etc

1.3. Data Privacy for data viewing

1.3.1. public designation of experiments allow for viewing of data by external to NIH members and internal NCATS members

1.3.2. private designation of experiments to allow for viewing of data by internal to NCATS members only

1.3.3. public or private designation should be available for individual experiments or an entire project folder

Module E – Creation of New Database and Migration of Historical data

1. Creation of a database specifically designed to store all future NCATS screening data.

This new database must be structured in a format to facilitate:

1.1. Raw screening data storage

1.2. Processed screening data storage

1.3. Compound information such as names and annotations

1.4. Compound plate maps

1.5. Assay metadata such as:

1.5.1. protein or cell line used

1.5.2. assay readout type

1.5.3. instrument used for data acquisitions

1.5.4. assay stage within the project

1.6. Dose response curve metrics generated from processed screening data

1.7. Drug combination screening metrics generated from processed screening data

2. Transfer historical data from existing NCATS databases to the new database, ensuring that all historical data is retained in a format which retains all critical information

3. New database must have the following technical capabilities:

3.1. Ability to connect with the application in Module B which will push data to the new database

3.2. Ability to connect with the applications in Module C and D which will have read access to the new database

3.3. Query of all stored data by API call to allow data to be accessible by various existing and future NCATS applications including by not limited to those existing in Module F

Module F – Integration with NCATS systems

1. Development and integration with the NCATS Open Data portal publicly accessible portal for viewing

1.1.1. Single and combination screen results,

1.1.2. analysis,

1.1.3. heat maps,

1.1.4. protocol information, and

1.1.5. compound information

2. Allow for results and promising compounds selected within this interface to also be combined with compounds identified in other areas of the NCATS OpenData portal

2.1. Unified application for exploration of this data such that results from these combination screenings can easily be compared to promising compounds identified in other modalities.

3. Provide a uniform look/feel to screening data in the NCATS OpenData portal such that user exploration of data is intuitive across data types.

4. Enable data to be downloaded for analysis by other collaborators

5. Integration with NCATS open collaboration tools including:

5.1. Developing all APIs with UNA authorization and authentication

5.2. Managing user access with NCATS UNA

3.0 REQUIREMENTS

3.1 TASKS

The following tasks outline the main task areas required for successful completion of Modules A-F:

1. Design, develop, and implement the NCATS screening data processing platform applications

a. Design and develop the platform

b. Test, market, and launch the new platform

c. Handover to operations and maintenance

3.2 DELIVERABLES

1) Application code artifacts for each Module

2) Deployment artifacts for each Module

3) Support for transition from development environment to production environment

4) Monthly progress reports

5) Final completion report