Computational Systems Biology Deep Learning in the Life Sciences - PowerPoint PPT Presentation

Computational Systems Biology Deep Learning in the Life Sciences 6.802 6.874 20.390 20.490 HST.506 David Gifford Lecture 20 April 21, 2020 COVID-19 Machine Learning Designed Therapeutics http://mit6874.github.io 1

Overview of today’s lecture COVID-19 and SARS-CoV-2 overview • COVID-19 epidemiology • COVID-19 testing • Vaccines for COVID-19 • Antibody therapeutics for COVID-19 •

Today’s deep learning methods • Vaccine design • Antibody discovery and improvement

COVID-19 (the disease) SARS-CoV-2 (the virus)

By NIAID - https://www.flickr.com/photos/niaid/49534865371

The basic reproduction number R 0 describes number of secondary infections from one individual https://en.wikipedia.org/wiki/Basic_reproduction_number

ACE2 bound to the 2019-nCoV S ectodomain with ~15 nM affinity, which is ~10- to 20-fold higher than ACE2 binding to SARS-CoV S. 10.1126/science.abb2507

This first preliminary description of outcomes among patients with COVID-19 in the United States indicates that fatality was highest in persons aged ≥85, ranging from 10% to 27%, followed by 3% to 11% among persons aged 65–84 years, 1% to 3% among persons aged 55-64 years, <1% among persons aged 20–54 years, and no fatalities among persons aged ≤19 years.

SARS-CoV-2 is a positive-sense single-stranded RNA virus that causes COVID-19 50 – 200 nanometers https://en.wikipedia.org/wiki/Coronavirus_disease_2019

SARS-CoV-2 is 29,903 bases and encodes 4 structural proteins (spike, envelope, membrane, nucleocapsid) RNA dependent RNA polymerase ORF1a and ORF1b (Remdesivir target) https://en.wikipedia.org/wiki/Severe_acute_respiratory_syndrome_coronavirus_2

Viral genome data suggests SARS-CoV-2 came from an animal https://doi.org/10.1038/s41591-020-0820-9

The receptor binding domain (RBD) of the spike protein is a primary therapeutic target https://doi.org/10.1038/s41586-020-2179-y

The spike protein (S) trimer ”up” component interacts with ACE2 on host cells

Interaction between RBD of spike and ACE2 ACE2 ACE2 RBD spike ACE2 RBD Lan et al., Nature March 30, 2020

Ma Massachusetts Consortium on on Pa Pathogen Readiness

Boston March 2, 2020

Guangzhou Institute of Respiratory Health (GIRH) Zhong Nanshan, Director

COVID-19 testing

Real-time quantitative PCR (RT qPCR) is used with primers specific to SARS-CoV-2 CT is the number of cycles required for a specific sample to cross the detection threshold https://www.ncbi.nlm.nih.gov/probe/docs/techqpcr/

Doubling time (R0 estimation) when testing is being introduced simultaneous To monitor and model a novel to epidemic escalation will obscure true epidemic growth pandemic, testing needs to be developed fast Doubling time estimates can be off and estimated cases off by orders of magnitude PCR is a good tool to do this True cases ? Detected cases time Testing capacity time Slide courtesy Michael Mina, Harvard

Prolonged viral shedding from multiple sites in severely ill patients A: Nasal swab B: Pharyngeal swab C: Sputum P T 1 P T 1 P T 1 P T 2 P T 2 P T 2 P T 3 P T 3 P T 3 P T 4 P T 4 P T 4 P T 5 P T 5 P T 5 P T 6 P T 6 P T 6 P T 7 P T 7 P T 7 P T 8 P T 8 P T 8 P T 9 P T 9 P T 9 P T 1 0 P T 1 0 P T 10 P T 1 1 P T 1 1 P T 11 P T 1 2 P T 1 2 P T 12 P T 1 3 P T 13 P T 1 3 P T 1 4 P T 1 4 P T 14 P T 1 5 P T 15 P T 15 P T 1 6 P T 16 P T 16 P T 1 7 P T 17 P T 17 P T 1 8 P T 18 P T 18 P T 1 9 P T 1 9 P T 19 P T 2 0 P T 20 P T 2 0 P T 2 1 P T 21 P T 2 1 P T 2 2 P T 22 P T 2 2 P T 2 3 P T 2 3 P T 23 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 19 2 0 2 1 2 2 2 3 2 4 25 2 6 2 7 28 2 9 3 0 31 3 2 3 3 34 3 5 3 6 37 3 8 D a y s p o s t-o n s e t D a y s p o s t-o n s e t D a y s p o s t-o n s e t D: Feces E: Urine F: Blood P T 1 P T 2 P T 1 P T 1 P T 3 P T 2 P T 2 P T 4 P T 3 P T 3 P T 4 P T 5 P T 4 P T 6 P T 5 P T 5 P T 7 P T 6 P T 6 P T 8 P T 7 P T 7 P T 9 P T 8 P T 8 P T 1 0 P T 9 P T 9 P T 1 1 P T 10 P T 10 P T 1 2 P T 11 P T 11 P T 1 3 P T 12 P T 12 P T 1 4 P T 13 P T 13 P T 1 5 P T 14 P T 14 P T 1 6 P T 1 5 P T 15 P T 1 7 P T 1 6 P T 16 P T 1 8 P T 17 P T 1 7 P T 1 9 P T 1 8 P T 18 P T 2 0 P T 1 9 P T 19 P T 2 1 P T 2 0 P T 20 P T 2 2 P T 21 P T 2 1 P T 2 3 P T 22 P T 22 P T 23 P T 23 1 2 3 4 5 6 7 8 9 1 0 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 1 9 2 0 2 1 2 2 2 3 2 4 2 5 2 6 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 3 6 3 7 3 8 3 9 4 0 4 1 4 2 1 2 3 4 5 6 7 8 9 10 1 1 1 2 1 3 1 4 1 5 1 6 1 7 1 8 19 20 21 22 23 24 25 26 2 7 2 8 2 9 3 0 3 1 3 2 3 3 3 4 3 5 36 37 38 39 40 41 42 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 D a y s p o s t-o n s e t D a y s p o s t-o n s e t D a y s p o s t-o n s e t PCR cycles to reach positive threshold Red = Positive Gray =Negative Ct value 16 18 20 22 24 26 28 30 32 34 36 38 40 Jincun Zhao, Guangzhou Institute of Respiratory Health

Antibody isotypes Affinity matured antibody specific to target Enhance phagocytosis of bound pathogens by macrophages Can cause antibody-dependent cell-mediated cytotoxicity (ADCC) Secreted antibodies – gut, mucus, tears, saliva, milk Can agglutinate pathogens to enhance their clearance Low affinity antibodies that are expressed early Activate the innate immune system Can agglutinate pathogens to enhance their clearance

MGH/Ragon ELISA Easy to Produce part of the Spike Full Spike Protein G1 full S G RBD 10 7 10 7 MFI full S COV2 10 6 MFI RBD COV2 10 6 IgG 10 5 10 5 10 4 10 4 10 3 10 3 + - CR3022 EBOV + - CR3022 EBOV COVID COVID A RBD A full S 10 6 10 6 MFI RBD COV2 MFI full S COV2 10 5 10 5 IgA Y 10 4 Y Wash away Y Addition of Y Addition of 10 4 Y 2 o antibodies excess antibodies Y substrate Y Y Y Y 10 3 10 3 + - CR3022 EBOV + - CR3022 EBOV Y Y Y COVID Y Y COVID Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y Y M full S M RBD 10 7 10 7 MFI full S COV2 MFI RBD COV2 10 6 10 6 Detection of IgG, IgA, or IgM IgM Receptor Binding Domain (RBD) 10 5 10 5 of the CoV2 Spike (S) 10 4 10 4 10 3 10 3 + - CR3022 EBOV + - CR3022 EBOV COVID COVID Slide courtesy Galit Alter, Ragon Institute

Sensitivity and unusual immune patterns Kinetics of response IgA1 IgM IgG1 ● 2000000 ● ● 750000 ● ● ● ● ● ● IgA 1500000 ● 2e+06 ● ● ● ● ● 500000 ● ● 1000000 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 1e+06 ● ● ● ● ● ● 250000 ● ● ● ● ● ● ● ● ● ● 500000 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0 ● ● ● ● ● 0e+00 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 5 10 15 20 5 10 15 20 5 10 15 20 days after symptoms days after symptoms days after symptoms IgA1 Defining accuracy days after symptoms days after symptoms IgA1 IgM IgG1 1.0 1.0 1.0 ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● ● 0.8 0.8 0.8 ● ● ● ● AUC AUC AUC ● 0.6 0.6 0.6 ● ● ● ● ~100% accuracy ~100% accuracy 0.4 0.4 0.4 ~100% accuracy 0 − 4 5 − 6 7 − 10 11 12 13 − 14 15 − 21 21 0 − 4 5 − 6 7 − 10 11 12 13 − 14 15 − 21 0 − 4 5 − 6 7 − 10 11 12 13 − 14 15 − 21 days after symptoms days after symptoms days after symptoms FcR2a FcR2b COV2_S COV2_RBD ● ● Slide courtesy Galit Alter, Ragon Institute

Mild patients have lower IgM responses against SARS-CoV-2 P T 1 P T 2 P T 3 P T 4 P T 1 3 P T 1 4 P T 1 5 P T 1 6 P T 5 P T 6 P T 7 P T 8 P T 1 7 P T 1 8 P T 1 9 P T 2 0 A P T 9 P T 1 0 P T 1 1 P T 1 2 P T 2 1 P T 2 2 P T 2 3 Controls IgM Severe IgM Mild IgM H ealth d o no r IgM S ev e re p a tie n t Ig M M ild patient IgM 1 .5 3 .5 3 .5 3 .0 3 .0 2 .5 2 .5 1 .0 IgM O D 450nm 2 .0 2 .0 1 .5 1 .5 0 .5 1 .0 1 .0 0 .5 0 .5 0 .0 0 .0 0 .0 H D P C N C 0 3 6 9 1 2 1 5 1 8 2 1 2 4 2 7 0 3 6 9 1 2 1 5 1 8 2 1 2 4 2 7 3 0 3 3 3 6 3 9 4 2 D a y s p o s t-o n s e t D a y s p o s t-o n s e t D a y s p o s t-o n s e t B Controls IgG Mild IgG Severe IgG H ea lth d o n o r Ig G S e v e re p a tie n t Ig G M ild patient IgG 1 .5 3 .5 3 .5 3 .0 3 .0 IgG 2 .5 2 .5 1 .0 O D 450nm 2 .0 2 .0 1 .5 1 .5 0 .5 1 .0 1 .0 0 .5 0 .5 0 .0 0 .0 0 .0 H D P C N C 0 3 6 9 1 2 1 5 1 8 2 1 2 4 2 7 3 0 3 3 3 6 3 9 4 2 0 3 6 9 1 2 1 5 1 8 2 1 2 4 2 7 D a y s p o s t-o n s e t D a y s p o s t-o n s e t D a y s p o s t-o n s e t Jincun Zhao, Guangzhou Institute of Respiratory Health