Becton, Dickinson and Company · NYSE:BDX
Becton, Dickinson and — Investor Presentation (2020)
53 slides from Becton, Dickinson and's investor presentation, 2020
Slide 1: BD Rhapsody Single-Cell Analysis System
BD RhapsodyTM Single-Cell Analysis System A robust microwell-based single-cell partitioning system for high-dimensional biology research 1
Slide 2: A complete single-cell multiomics offering
A complete single-cell multiomics offering BD RhapsodyTM BD RhapsodyTM BD RhapsodyTM Analysis Single-Cell Analysis System Assays & Reagents Pipelines & SeqGeqTM Software Whole Transcriptome Amplification WTA BD RhapsodyTM BIOINFORMATICS Targeted Targeted RNA-Seq WTA and INSTRUMENT Targeted REAGENTS
Slide 3: BD Rhapsody Systems – two configurations
BD Rhapsody™ Systems – two configurations BD Rhapsody™ Single-Cell Analysis System BD Rhapsody™ Express Single-Cell Analysis System • BD Rhapsody™ Scanner and BD Rhapsody™ Express • BD Rhapsody™ Express System System • The same powerful single-cell partitioning • Visual workflow QC system without vi
Slide 4: BD Rhapsody System workflow
BD Rhapsody™ System workflow Does not use microfluidics • Cells settle into microwells by gravity • Volume of microwell array is ~600 µL, cell suspension loaded can be quite dilute • Clogging of channels with cells is less of a concern Requires minimal set of benchtop equipment • Fluidic exchange in
Slide 5: Unique benefits of the BD Rhapsody System
Unique benefits of the BD Rhapsody™ System Microwell-based single-cell partitioning Visual workflow QC Magnetic Univ CL UMI Poly-dT Bead Stable bead-oligo complex 5
Slide 6: Microwell-based single-cell partitioning
Microwell-based single-cell partitioning Trust the BD Rhapsody™ System with your precious samples Superior Cell Capture Rates Low Multiplet Rates Highest Cell Throughput per Flow Cell Go Beyond PBMCs Reliable, Consistent Results 6
Slide 7: Superior cell capture rate
Superior cell capture rate Across 75 runs, ~75% of cells loaded onto • ~5000 Jurkat/Ramos cells (1:1 mix) cartridge are retrieved by sequencing • Experiments performed on multiple days by multiple users Overall Intrasample Mean Median variability variability • Data taken from a total of 75 (SD, all
Slide 8: Cartridge capture rate for various cell types
Cartridge capture rate for various cell types Live cells loaded to Cartridge Capture Cell type Tissue Sample prep cartridge* rate** CAR-T cells Blood Cell in-house manufacturing 25,000 72% Frozen, ON grown and trypsinised Mesenchymal stem cells (MSC) Fatty tissue cells 25,000 73% Dissociated tumor o
Slide 9: Examining capture of cell types compared to flow cytometry
Examining capture of cell types in the BD Rhapsody™ System compared to flow cytometry Bone marrow extracted 9
Slide 10: High correlation between cell types seen with BD Rhapsody and flow cytometry
High correlation between cell types seen with the BD Rhapsody™ System and those seen using flow cytometry % of live cells Mouse 1 % of live cells Mouse 2 10
Slide 11: Faithful capture of cells with disparate size and morphology
Faithful capture of cells with disparate size and morphology 5000 Cartridge 1 4500 4000 # of cells 3500 Cells Loaded Cartridge 2 3000 2500 Cells Retreived by Sequencing: Replicate 1 2.5:1:1:1 Neutrophils:Jurkat:K562:BT456 2000 Replicate 2 1500 1000 Neutrophils* (10-14 um) 500 Jurkat (11 um) 0 Neutro
Slide 12: High cell capture and low multiplet rate across cell inputs
High cell capture and low multiplet rate across cell inputs Theoretical (poisson) 14.0% Scanner 20,000 Cells 12.0% 10,000 Cells 40,000 Cells Multiplet rate 10.0% 8.0% Target cell Live cells loaded Viable cells captured Capture input (hemocytometer count) in wells with a bead rate 6.0% 4.0% *40,000 5
Slide 13: Reproducible data with minimal cell-cell crosstalk
Reproducible data with minimal cell-cell crosstalk from high cell inputs Displaying 3,000 cells from a 40,000 cell load (1:1 mixture of human and mouse cell lines; shallow sequencing) Multiplet Rate (n=3) 12.0% Replicate 1 Replicate 2 10.0% Human molecules/cell Human molecules/cell 8.0% 6.0% 4.0% 2.
Slide 14: Superloading enables significant cost savings
Superloading enables significant cost savings Load 40,000 cells 1 Sample Process 40,000 cells as one sample ~8% multiplet rate @ 40,000 cell load >220,000 partitions per flow cell Platform Assay Sample Input BD Rhapsody™ Cartridge + Reagents Cost BD Rhapsody™ Targeted 40,000 cells per cartridge w/ S
Slide 15: Go beyond PBMCs with the BD Rhapsody System
Go beyond PBMCs with the BD Rhapsody™ System • Increasing recognition in the Immunology field about the importance of granulocytes in immune response https://youtu.be/i7UIYdyFJvQ • Microwell-based systems – Gold standard for fragile cells such as granulocytes – Reliable – Much less batch effects all
Slide 16: Severe COVID-19 is marked by a dysregulated myeloid cell compartment
Severe COVID-19 is marked by a dysregulated myeloid cell compartment https://doi.org/10.1016/j.cell.2020.08.001 16
Slide 17: Single-cell characterization of human cortical spheroids
Single-cell characterization of human cortical spheroids 17
Slide 18: Identification of a distinct arthritis associated macrophage population, AtoM
Identification of a distinct arthritis associated macrophage population, AtoM 18
Slide 19: Comprehensive census of human meniscus cells using single-cell RNA-Seq
Comprehensive census of human meniscus cells using single-cell RNA-Seq 19
Slide 20: Reliable, consistent results
Reliable, consistent results • Minimal batch effects – Technical replicates – Biological replicates – Site to site – User to user • Minimum sample bias • Ability to integrate large data sets without correction for batch effects • System of choice for translational, multi-center, multiple cohort stud
Slide 21: T cell activation time-course examined across technical replicates
T cell activation time-course examined across technical replicates Percentage of cells assigned to each sample tag per cartridge 20 18 16 14 12 10 8 6 4 2 n=2 0 Cartridge 1 Cartridge 2 21
Slide 22: No batch effect from technical replicates
No batch effect from technical replicates 50 Cartridge Sample_Tag 1 40 tSNE projection: Cartridge 1 vs Cartridge 2 30 File name 20 50 10 SampleTag07_hs (18.2%) SampleTag08_hs (16.0%) Coord 2 40 SampleTag09_hs (13.8%) 0 SampleTag10_hs (16.0%) Cartridge 1 SampleTag11_hs (20.0%) 30 -10 Sample Tag 7 Sam
Slide 23: Strong correlation of mRNA and protein expression between cartridges
Strong correlation of mRNA and protein expression in resting and activated T cells between cartridges 40-plex BD® AbSeq Reagents All mRNA and protein markers Correlation: mean( log10(molecules per cell per gene) ) Correlatio Resting T cells 3 3 Cartridge 1|SampleTag07_hs| 2 R =0.996 Cartridge 2|Samp
Slide 24: Bone-marrow derived cells examined from biological replicates
Bone-marrow derived cells examined from biological replicates Cartridge 1 (Mouse 1) Bone marrow extracted Cartridge 2 (Mouse 2) n=2
Slide 25: Similar proportions of each cell type recovered from biological replicates
Similar proportions of each cell type recovered from biological replicates Mouse 1 Mouse 2 Mouse 2 Mouse 1 myeloid derived Mouse #1 supressor cells neutrophils Mouse #2 immature neutrophils stem cells monocyte B cells CD8 T cells
Slide 26: Site-to-site reproducibility examined by targeted gene and protein expression
Site-to-site reproducibility examined by targeted gene expression and protein expression analysis when comparing two sites Frozen PBMCs Site 1 Site 2 Single-cell Single-cell workflow workflow +targeted +targeted mRNA and mRNA and ab-oligos ab-oligos
Slide 27: Good correlation between mRNA and protein markers between sites
Good correlation between mRNA and protein markers between sites mRNA Protein (BD AbSeqTM Antibodies) Correlation: mean( log10(molecules per cell per gene) ) Correlation: Correlation: mean( log10(molecules mean( log10(molecules per cell per per gene) cell )per gene) ) Correlation: 1.5 1.5 2.5 2.5 BD
Slide 28: Little batch effect observed when comparing two sites
Little batch effect observed even with no batch correction applied when comparing two sites Replicates annotation Immune Panel mRNA + 22 AbSeq 50 50 40 40 30 30 20 20 B CD4T_memory 10 10 CD4T_naive CD8T_memory Coord 2 Coord 2 Replicate 1 CD8T_naive 0 0 Replicate 2 DC Monocyte_classical Monocyte_nonc
Slide 29: High correlation between data generated by three different operators
High correlation between data generated by three different operators • BD RhapsodyTM Human Immune Response Panel • 1:1 mix of 5,000 Jurkat and Ramos cells were loaded onto each cartridge 29
Slide 30: Consistent results with minimum sample bias across different cell inputs
Consistent results with minimum sample bias across different cell inputs Workflow using the tSNE colored by cell input BD Rhapsody™ System 100 cells 500 cells Sequencing and analysis PBMCs 1,000 cells 5,000 cells 10,000 cells 30
Slide 31: Similar proportions of each cell type recovered from different cell inputs
Similar proportions of each cell type recovered from different cell inputs Percent of cell type in each library 50 45 40 35 Percentage 30 25 20 15 10 5 0 100 500 1k 5k 10k 31
Slide 32: Consistent performance with pre-formed microwell partitions
Consistent performance with pre-formed microwell partitions Flow Profile and Leak Testing Functional Testing Physical Attribute Testing Technical replicates User-to-user Biological replicates Site-to-site R2 = 0.995 R2 = 0.963 User 2 R2 = 0.997 R2 = 0.973 Replicate 2 R2 = 0.973 Mouse 1 Site A User 3
Slide 33: Stable bead-oligo complex
Stable bead-oligo complex Peace of mind and efficiency Similar Data from Subsampled Beads Similar Data from Archived Beads Iterate on Experiment Design Tool to Standardize Performance 33
Slide 34: Expected number of cells seen after bead subsampling
Expected number of cells seen after bead subsampling 20,000 Split beads: PBMCs Beads • 100 µL • 50 µL • 25 µL • 12.5 µL • 2x6.25 µL 8000 Cells seen in sequencing 7000 6000 5000 4000 3000 2000 1000 0 0 2000 4000 6000 Cells expected based on subsampling 34
Slide 35: High correlation and no batch effect observed with bead subsampling
High correlation and no batch effect observed with bead subsampling 35
Slide 36: Equivalent data obtained from stored beads
Equivalent data obtained from stored beads PBMCs Beads Split beads: • Fresh • 6 weeks (4C) • 12 weeks (4C) • No batch effect between beads stored for 12 weeks • Good correlation of gene expression between data from fresh beads to beads stored for 12 weeks 36
Slide 37: Focus on strengths, leverage partners
Focus on strengths, leverage partners • Generate beads with stable cDNA that can be stored for up to 3 months • Perform single-cell capture to RT and Exo treatment steps and store beads. Collect all clinical or research samples to process further, so all libraries can be processed on same day to red
Slide 38: Iterate on experiment design
Iterate on experiment design • Perform multiple assays over time on the beads prepared from the same set of single cells • Targeted mRNA, AbSeq, and whole transcriptome libraries can be generated from the same stock of oligo-bead complex • Allows for recovery from experimental failures • Gain confid
Slide 39: Similar proportions recovered from WTA and Targeted assays on subsampled beads
Similar proportions of each cell type are recovered from WTA and Targeted assays performed on subsampled beads 35 30 25 20 % Cells 15 10 5 0 Targeted WTA B Cells CD4 Memory T Cells CD4 Naïve T Cells CD8 Memory T Cells CD8 Naïve T Cells Dendritic Cells Monocytes (classical) Monocytes (nonclassical) N
Slide 40: Tool to standardize performance
Tool to standardize performance • Stored, subsampled beads as a tool to: – Test technical skills of users • New user vs Experienced user • Users across different laboratories • CROs and collaborators – Develop SOPs – To control for day-to-day variability in downstream library prep 40
Slide 41: Scanner for workflow QC
Scanner for workflow QC Not flying blind, confidence with every experiment Visual Workflow QC with Analysis Metrics at Every Step Estimate Cells Expected from Sequencing See Before You Pay Troubleshooting and Optimization of Single-Cell Capture without Sequencing 41
Slide 42: Visual workflow QC
Visual workflow QC Pairing cells with beads in microwells: Load cells … … Array with > 200,000 microwells 42
Slide 43: Bead loading
Bead loading Pairing cells with beads in microwells: Load beads Microwell Barcoded Bead Cell … … 43
Slide 44: Cell lysis
Cell lysis Lyse cells to release mRNA, which hybridizes to label molecules on bead mRNA from cell Barcoded Bead Lysed cell … … 44
Slide 45: Bead retrieval
Bead retrieval Retrieve beads and pool to a single tube … … … 45
Slide 46: Analysis metrics
Analysis metrics • Compare cell multiplet rates by imaging vs calculated by Poisson distribution • Make Go/No-Go decisions • Estimate cells retrieved by sequencing 46
Slide 47: Analysis metrics at every step
Analysis metrics at every step Scanner metric Use Number of wells with viable cells at cell load Provides a preliminary estimate of the number of wells with viable cells captured Provides a preliminary estimate of the total number of viable cells captured in the Number of viable cells captured in we
Slide 48: Scanner metrics are highly correlated with recovery by sequencing
Scanner metrics are highly correlated with recovery by sequencing, helping users better plan for sequencing cost R h a p so d y st a c ke d yie ld C e ll lo a d t o se q y ie ld 5500 Hemocytometer-based [# rse c c e lls] / [live c e lls lo a d e d t o ccell a rt r idcounts ge] Cartridge-based cell c
Slide 49: See before you pay!
See before you pay! Live cells loaded # wells with viable Total number of reads Sequencing cost cells and a bead needed (assuming targeted + 10-plex BD® AbSeq Reagents) 20,000 cells 20,000 cells 233M NextSeq® System High Output ($2,600, ~400M reads) 20,000 cells 9,000 cells 105M NextSeq® System Mid
Slide 50: Troubleshooting without scanner is tedious
Troubleshooting without scanner is tedious • Troubleshoot cartridge loading – Multiplet rate higher than expected - may indicate cell clumping – Cell viability and total number of cells captured in cartridge, after cell load vs in cell prep – indicates if a significant number of cells initially load
Slide 51: Use scanner to optimize experimental workflow to maximize cell recovery
Use scanner to optimize experimental workflow to maximize cell recovery without paying for sequencing BD Rhapsody™ System Upstream sample prep that may affect cell capture workflow optimization Fresh sample Frozen sample % cells captured on cartridge Lysis 1 Lysis 2 Lysis 1 Lysis 2 Settling time 51
Slide 52: BD Rhapsody System - A robust single-cell partitioning system with unique features
BD Rhapsody™ System - A robust single-cell partitioning system with unique features Microwell-based single-cell partitioning Visual workflow QC Magnetic Poly-dT Univ CL UMI Bead Stable bead-oligo complex 52
Slide 53: Thank you!
Thank you! Class 1 Laser Product. For Research Use Only. Not for use in diagnostic or therapeutic procedures. 23-23087-00 BD Life Sciences, San Jose, CA 05131, USA bdbiosciences.com BD, the BD Logo, FACSLyric, Rhapsody and SeqGeq are trademarks of Becton, Dickinson and Company or its affiliates. All
Becton, Dickinson and Company — Investor Presentation (2020) on 6ix