Profiling how Immune inhibitors Secreted by Melanoma affect NK & other immune cells

| 0

Presented By

Date

Profiling how Immune inhibitors Secreted by Melanoma

affect NK & other immune cellsDiscovery On Target, Boston, Sept 26th

Anton Yuryev

| 1

Gaining insights from internal and external/public

information sources is time and resource consuming

We have over 500 information solutions, we

don’t want 501. We want to consolidate our

solutions and increase information

discoverability.Pers. Comm. Head of Medicinal

Chemistry at aTop5 Pharma

The challenge is in putting together different data sources and seeing patterns.

Former Pharma COO

There is lots of locked away data—if that could be made available, it would be highly valuable.

CIO of Biotech

64 % of data management effort and time

is spent finding and profiling data sources

55-75 % of data collected

by businesses go unused

Unstructured

data

Structured

data

Source: Forrester Research Survey, Global Databerg Report

| 2

Gain critical insights from current,

integrated, knowledge helps to better

inform your critical workflows: Identify novel immunotherapy targets

Find potentially new immunomodulatory

drugs through drug repurposing

Create models that improve

understanding of combinatorial

treatment interactions

Better match patients with treatments

Discussion Summary

| 3

Elsevier’s Solutions can help streamline critical workflows with

integrated data

Discovery & Lead ID & Valid Pre-clinical Clinical Post-launch

• Monitoring adverse

events

• Lead prioritization for safety, delivery and efficacy

• Translational medicine/research

• Lead identification and

characterization

• Synthesis optimization

• Bioactivity

• Disease modeling

• Target identification

• Biomarker discovery

• Drug repositioning

Elsevier and non-Elsevier

textual information

Public and proprietary

databasesDisparate Data/Content

Examples

Supported

Applications

Use-case centered integration & customization focus on customer outcomes

Expertise/ Capabilities Data extraction, Data normalization, Data integration

Elsevier Text MiningTechnology & data structure Dictionaries & taxonomies

Outcomes

+

| 4

• Human γδ T cells lyse melanomas and other cancerous epithelial

cells in a perforin-mediated manner.

• Indeed, melanoma cell lines are subject to lysis by γδ T-cells,

which produce perforin and exhibit strong cytolytic activity upon

exposure.

• Both in vitro and in xenograft models, γδ lymphocyte-mediated

cytotoxicity against melanoma cells has been reported.

• Our results suggest that a natural immune response mediated by

γδ T lymphocytes may contribute to the immunosurveillance of

melanoma.

• Killer cell inhibitory receptors for MHC class I molecules regulate

lysis of melanoma cells mediated by gamma delta T

lymphocytes.

Transform text to facts using Elsevier Deep Reading technology

gamma delta T cell melanoma

negative

regulation

| 5

Find the

right data

NLP information extraction data model: 201746 relation types; >7,138,122 relations; >32,658,982 facts

Relation Type Stats

Binding 324,913

ProtModification 57,830

DirectRegulation 38,562

PromoterBinding 34,104

microRNAEffect 39,181

MolTransport 40,515

Expression 286,648

Regulation 382,111

Type Stats

Binding 75,911

MolSynthesis 24,688

MolTransport 30,222

ChemReaction 76,744

Expression 340,815

MolTransport 11,736

DirectRegulation 87,312

Regulation 413,152

Type Stats

Regulation 826,471

Regulation 438,389

Expression 16,069

MolTransport 1,730

MolSynthesis 16,102

MolTransport 5,618

Clinicaltrial 4,067

Regulation 127,515 Protein->Clinical

parameter

Type Stats

Regulation 616,577

Regulation 566,444

ClinicalTrial 78,713

QuantitativeChange 250,475

GeneticChange 193,858

StateChange 16,291

Biomarker 73,677

QuantitativeChange 29,300

Biomarker 9,025

FunctionalAssociation 410,856

FunctionalAssociation 169,586

Type Stats

FunctionalAssociation 5,062

Regulation 13,854

Regulation 57,179

MolTransport 15,834

CellExpression 432,591

MolTransport 51,908

Biomarker 4,016

QuantitativeChange 8,112

StateChange 9,328

Regulation 102,435

Regulation 135,789

CellEffectTM

for cancer

immuno-

therapy

| 6

Epitope

Normalization of cell identifiers:

From inconsistent names to standard names

Basic cell“Attribute”

CD4+ CD25+ regulatory T cell

T-lymphocyte leukocyte

T-cell leucocyte

hemopoetic

hemopoietic

haemopoetic

haemopoietic

hematopoetic

hematopoietic

haematopoetic

haematopoietic

regulatory

immunoregulatoryCD4+CD25+

CD25+FOXP3+

CD4+ CD25+ FOXP3+

CD3+CD4+CD25+

Standard

cell name

| 7

Example of a “synonym” list

| 8

Provide greater context by adding cell processes to

each cell type

Allows more flexibility to address emerging topics • More cell processes in the database provides greater breadth of information

• Find cell processes relevant to rare and cell types critical to biology of disease

• Automatic tracking of changes in literature trends to keep pace with evolving biology

proliferation of

death of

migration of human polarization

cytotoxicity

quantity

Standard

cell

name

| 9

Modeling to identify immunotherapy melanoma targets in

Pathway Studio – graphical query in Pathway Studio

May help some cancers to grow

| 10

Expanding the query:

Melanoma-related concepts in Pathway Studio

178 melanoma

cell lines

32 melanoma

Disease concepts

| 11

Expanding the query:

Concepts related to immune suppression

291 concepts that can be

positively regulated to

suppress immune response

1376 concepts that can be

negatively regulated to

suppress immune response

| 12

Known melanoma Immune suppression mechanisms• 226 proteins secreted or expressed on the cell surface by melanoma that can inhibit

activation of immune system

• 142 proteins secreted or expressed on the cell surface by melanoma that can activate

immune toleranceCellExpression MolTransport relation types

Regulation Effect=negative or positive relation types

1. Extracted from more than 20,000 articles and network created in several hours

2. Each target requires at least two publications:

describing its expression in melanoma

describing its immune system suppressive functionOnly NLP extraction allows search

across several articles

| 13

Main result of inspecting potential

immunotherapy targets:

Many mechanisms, not one.

Examples of targets found by curation of the results of

expanded queries

13

| 14

Example why Keytruda

may not work:

PDCD1 is not the only

mechanism activating

immune tolerance

Example of novel

target with no drugs

Examples of known and novel targets

| 15

Example of drug repurposing: Galectin-1; VEGFA/C

| 16

Well-known drug may be used in combinatorial immunotherapy with

Keytruda

CD73 –

Oleclumab target

| 17

More novel targets for immunotherapy

TEW-7197 target

| 18

Multiple immuno-modulatory mechanisms can help

identify new, potentially, safer drug combinations

| 19

Alternative mechanism for immuno-suppression in the tumor

and low PDCD1 expression in a patient should predict no

response to Keytruda

| 20

Find personalized immunotherapy option targeting a

mechanism activated in a single patient.

Precision Oncology 3.0

(2020)

| 21

Gain critical insights from current,

integrated, knowledge helps to better

inform your critical workflows: Identify novel immunotherapy targets

Find potentially new immunomodulatory

drugs through drug repurposing

Create models that improve

understanding of combinatorial

treatment interactions

Better match patients with treatments

Discussion Summary

| 22

www.elsevier.com/rd-solutions

Thank you for your attention.

Acknowledgements:

Maria Shkrob, PhD

Stephen Sharp, PhD

Mathew Clark, PhD