Chromium™  Single  Cell  3’SolutionHigh-­‐throughput  Single  Cell  Gene  Expression

July  2016,  Revision  B

2

The  Chromium  Single  Cell  3’  Solution

• GemCode™  Technology

• Automated

• High-­‐throughput   and  scalable

Chromium™ Controller

• Chip  for  single   cell  partitioning   in  GEMs

• Reagents   for  RT,  amplification   and  library  construction  

Chromium   Single  Cell   3’  Consumables

• Informatics   solution  for  single   cell  expression   profiling

• Pre-­‐processing,   QC  and  analytics

Cell   Ranger™   Pipelines

I

3

Gel  Bead-­‐in-­‐Emulsion   (GEM)

Single  T  Cell

Functionalized  Gel  Bead

RT  Reagents  in  Solution

Partitioning   Oil

4

High  Diversity  Barcode  Library

High-­‐diversity  Library

Functional  Oligo  with  BarcodeGel  Bead   Scaffold

R2P7 10X  BC Poly(dT)VN

• 750,000  Discrete   Reagents   in  One  Tube• Defined   14  base   pair  barcode  sequence• Highly  uniform  size  and  barcode  representation• Built-­‐in   sequencing   adapter,  barcode  and  primer

5

Partitioning  and  Barcoding  of  Single  Cells

• Super-­‐Poisson   loading   of  barcoded  beads  into  droplets• Poisson   loading  of  cells   in  GEMs• Beads   dissolve   for  efficient,   liquid   phase  biochemistry• Cell   lysis  starts   immediately   following  encapsulation

6

Efficient  Scalable  Workflow

Single-­‐use  microfluidics  chip

User  controlled  trade-­‐off   between   cell  numbers  and  doublet  rate

Gel  Bead  wellSample  well

Oil  well

Recovery  well

• Up  to  8  channels   processed   in  parallel• 1,000  to  6,000  cells   per  channel• 10  minute   run  time   per  chip• Up  to  30  um  cell  diameter   tested• ~50  %  cell   processing   efficiency• Temperature   Range   18-­‐280C

Number of  Cells Expected   Doublet Rate  (%)*

1,200 ~1.2

3,000 ~2.9

6,000 ~5.7

7

Assay  Scheme  for  3’  mRNA  Sequencing

In  GEM RT  with  oligo(dT)VNmRNA

GGG  

CCC

Template  Switching  RT

First  Strand  Synthesis

TSO

CCCOligo(dT)VNR210XP7

Oligo(dT)VNR210XP7

Oligo(dT)VNR210XP7

8

Assay  Scheme  for  3’  mRNA  Sequencing

cDNA  amplification  by  PCR

Shearing,  A  tailing,  ligation  and  SI  PCR

Primer

R1 SI P5

CCC

Primer

Oligo(dT)VNR210XP7

Oligo(dT)VNR210XP7

In  Bulk

9

Single  Cell  3’  End-­‐to-­‐End  Workflow

Cell   preparation1

Partition   and  RT  inside   each  GEM2

Pool  and  cDNA  amplification3

Covaris shearing4

Adapter   ligation   and  sample   index   PCR5

Sequencing   and  analysis6

Total  Turn-­‐around   Time:  ~12  HrsTotal  Hands-­‐on   Time:  ~4  Hrs

Reagents   and  Consumablesin  10X  Kit

10

• Complete  software  package  for  single  cell  analysis

• Bundled  with  STAR  for  efficient  transcriptome  alignment

• Outputs  are  standard  formats  plus  Loupe  visualization

Cell  Ranger  – Informatics  Workflow

Cell  Ranger™Analysis  Pipelines

Loupe™Cell  Browser

Standard  Informaticssamtools,  Python,   R

BAM

HDF5

MEX

LOUPE

BCL

11

Cell  Ranger  – Output  Files

• Indexed  BAM  containing  position-­‐sorted,  aligned  reads• Barcodes  and  UMIs  attached  as  standard  tags

BAM  – Genome-­‐Aligned  Reads

• Market  Exchange   format• Suitable  for  downstream  analysis  in  Python  and  R

MEX  – Gene/Barcode  Matrix

• Run  metrics  and  basic  static  visualizations

HTML,  CSV  – Run  Summary

CSV  -­‐ Run  Analysis

Gene ATCAGGGACAGA AGGGAAAATTGA TTGCCTTACGCG TGGCGAAGAGAT TACAATTAAGGC

LOXL4 0 0 0 0 0

PYROXD2 1 0 1 1 0

HPS1 23 12 9 8 3

CNNM1 0 2 1 0 0

GOT1 22 6 7 9 3

• 2D  projections• Cell  clustering• Differential   expression

12

Example  1:  Validation  of  Single  Cell  Behavior

Human:Mouse

Human  only

Mouse  only

Human  Transcript   Counts

Mou

se  Transcript  Co

unts

60,00000

• 1:1  mixture   of  ~1,400  human  (HEK293T)  and  mouse   (NIH3T3)  cells

• 99.4%  of  cell-­‐occupied  GEMs   yielded   reads  mapping   to  only  one  species

• 1%  inferred  doublet   rate*  

*includes  unobserved  human:human and  mouse:mouse doubletsNumber  of  cells  detected:   ~1400  cells,  Number  of  raw  reads  per  cell:  ~130k

13

Example  2:  Cell  Cycle  Phases

Combined   Expression   of  Known  Phase  MarkersG1/S

S

G2

G2/M

M/G1Expression

level

HEK293T  Cells   Ordered  by  Inferred  Cell   Cycle  Phase  

0 100 200 300 400

• Proliferating  HEK293T  cells  were  profiled   and  scored  for  expression   of  markers  associated   with  each  major  cell  cycle  phase

• Cells   from  all  phases  were  identified

Phase-­‐specific   genes  derived  from  Whitfield  et  al.,  2002Number  of  cells  detected:   ~400  cells,  Number  of  raw  reads  per  cell:  ~40k

14

Example  3:  Breast  Cancer  Heterogeneity

HCC1954

HCC38

HCC1143

t-­‐SNE  Projection

log(x+1)  HER2counts  

HCC1954

HCC38

HCC1143

t-­‐SNE  Projection

Unbiased   Automatic   Clustering   of  Three  Breast   Cell  Lines

HER2 Expression   Matches  Expected   Cell   Line  Status

Number  of  cells  detected:   ~1000  cells,  Number  of  raw  reads  per  cell:  ~40k

15

Example  4:  Identifying  Rare  Cell  TypesB-Cell (Raji) T-Cell (Jurkat)

-20 -10 0 10 20PC1

-20 -10 0 10 20PC1

-20 -10 0 10 20PC1

-20 -10 0 10 20PC1

-20 -10 0 10 20PC1

-20

-10

0

10

20

PC

2

-20

-10

0

10

20

PC

2

-20

-10

0

10

20

PC

2

-20

-10

0

10

20

PC

2

-20

-10

0

10

20

PC

2

12% 1.5% 0.6%

• Jurkat and  Raji cells  were  combined   at  9:1,  99:1  and  199:1  ratios  and  then  profiled

• The  minority   Raji populations  were  identified   in  all   three  mixtures

Number  of  cells  detected:   ~1000  cells,  Number  of  raw  reads  per  cell:  ~60k

16

Example  5:  Primary  Cell  Populations

Platelets  andPlasma  Components

Erythrocytes(RBCs),  Eosinophilsand  Neutrophils  

Myeloid  Cells Lymphoid  Cells

Monocytes

MyeloidDendritic  Cells

Whole  Blood

Macrophages

Plasmacytoid  DendriticCells

B  Cells NK  Cells T  Cells

Mononuclear  Cells(PBMCs)• A  complex  mixture  of  different  cell  

types• Well-­‐studied   and  readily  available  primary  cells

Peripheral  Blood  Mononuclear  Cells  (PBMC)

17

Major  populations  of  PBMCs  are  detected

CD4+  T  (28.4%)

CD14+  Monocytes  (5.3%)

CD56+  NK  (13.5%)

CD34+  Progenitors  (0.3%)

Dendritic  (1.9%)

CD19+  B  (5.5%)

CD8+  T  (18.7%)

CD45  RA+  Naïve  T  (26.4%)

TSNE1

TSNE2

18

Example:  68,000  Human  PBMCs

• CD45RA+Naïve   T  Cells• CD4+  T  Cells• CD8+  T  Cells• CD14+  Monocytes• CD19+  B  Cells• CD34+  Myeloid   Progenitors• CD56+  Natural   Killer  Cells