Method for determination and identification of cell signatures and cell markers

• US 11,225,689 B2
• Filed: 08/17/2017
• Issued: 01/18/2022
• Est. Priority Date: 08/17/2016
• Status: Active Grant

• Current Assignee: President and Fellows of Harvard College (Harvard Corporation), Howard Hughes Medical Institute, Massachusetts Institute of Technology, The Broad Institute, Inc.
• Original Assignee: The Broad Institute, Inc.
• Inventors: Shekhar, Karthik, Lapan, Sylvain, Whitney, Irene, Macosko, Evan, McCarroll, Steven, Cepko, Constance, Regev, Aviv, Sanes, Joshua
• Primary Examiner(s): Negin, Russell S
• Application Number:
  US15/680,127
• Publication Number:
  US 20180127823A1
• Time in Patent Office: 1,615 Days
• CPC Class Codes: C12N 5/0621 Eye cells, e.g. cornea, iris pigmented cells photoreceptors C12N5/062
• CPC Class Codes: C12Q 1/6841 In situ hybridisation
• CPC Class Codes: C12Q 1/6881 for tissue or cell typing, e.g. human leukocyte antigen [HLA] probes
• CPC Class Codes: C12Q 2600/158 Expression markers
• CPC Class Codes: G01N 2800/164 Retinal disorders, e.g. retinopathy
• CPC Class Codes: G01N 33/56966 Animal cells
• CPC Class Codes: G16B 25/00 ICT specially adapted for hybridisation; ICT specially adapted for gene or protein expression
• CPC Class Codes: G16B 25/10 Gene or protein expression profiling; Expression-ratio estimation or normalisation
• CPC Class Codes: G16B 40/00 ICT specially adapted for biostatistics; ICT specially adapted for bioinformatics-related machine learning or data mining, e.g. knowledge discovery or pattern finding
• CPC Class Codes: G16B 40/20 Supervised data analysis
• CPC Class Codes: G16B 40/30 Unsupervised data analysis

Abstract

The present invention provides for methods of identifying cell types and cell subtypes from a biological sample or population of target cells. The methods further provide for determining cell type or cell subtype signatures. The method further provides for bipolar cell subtypes and markers and cell signatures thereof.

Claims

1. A method of identifying a cell or cell marker, comprising

a) isolating target cells by label-free imaging flow cytometry;

b) quantifying gene expression in the target cells by single cell sequencing,

wherein sequenced transcript reads are obtained from RNA sequencing of 20,000 or more single cells from a biological sample, wherein the sequencing depth is about 10,000 or less reads per cell;

c) clustering the target cells based on the gene expression by application of one or more algorithms;

d) determining a transcription signature for each cluster based at least in part on identifying differentially expressed genes between two or more clusters and between each cluster and the remaining cells as background by applying an unsupervised machine-learning model; and

e) validating gene expression against cellular morphology.

2. The method of claim 1, wherein the transcription signature identifies a bipolar cell type.

3. The method of claim 2, wherein the transcription signature identifies a rod bipolar cell, or one of the following cone bipolar cells:

BC1A, BC1B, BC2, BC3A, BC3B, BC4, BC5A, BC5B, BC5C, BC5D (XBC), BC6, BC7, BC8, and BC9.

4. The method of claim 3, wherein the transcription signature identifies a bipolar cell and comprises at least one of Pcdh17, Pcdh10, Erbb4, Nnat, Col11a1, Sox6, Chrm2, Slitrk5, Lrrtm1, Cck, Lect1, Igfn1, Serpini1, Cpne9, Vstm2b, Casp7, Vsx1, Vsx2, Otx2, Scgn, Slcla2, Slc6a9, Pax6, Tfap2a, Stx1b, Tacr3, Bhlhe23, Fat4, Fezf1, Fezf2, Wfdc2, Mylk, Hs3st2, Wls, Nxph1, Lhx3, Cabp5, Grm6, Kcng4, Cdh9, Htr3a, Hcn1, Hcn4, Nfia, Cntn5, BC046251, Gfra3, Areg, Kirrel3, Isl1, Spock3, Serpini1, Prkca, Car8, Sebox, Apoe, GluI, Aqp4, Cabp5, Syt2, Grik1, Prkar2b, Rcvrrn, Irx6, Gad1, Rho, Pdc, Nr1, Pde6a, Pde6h, Arr3, Opn1mw, and Opn1sw.

5. The method of claim 1, wherein validating gene expression against cellular morphology comprises sparsely labeling the cell to enhance the expression of a fluorescent protein in the cell and combining the sparse labeling with fluorescent in situ hybridization (FISH) to validate the marker against cellular morphology in step e).

6. The method of claim 5, wherein FISH is combined with a specific antibody, double FISH, or a transgenic reporter mouse line directed to a previously identified marker in the cell.

7. The method of claim 1, wherein the unsupervised machine-learning model is a k-means model, a mixture model, a hierarchical clustering model, an anomaly detection model, a neural network model, an expectation-maximization (EM) model, a method of moments model, or a blind signal separation technique.

8. A method of determining transcription signatures for cell types or cell sub-types, comprising:

isolating a population of cells by label-free imaging flow cytometry;

identifying cell types and/or cell sub-types by applying one or more clustering algorithms to a set of sequenced transcript reads from the population of cells,

wherein the sequenced transcript reads are obtained from RNA sequencing of 20,000 or more single cells from a biological sample, wherein the sequencing depth is about 10,000 or less reads per cell,

wherein each resulting cluster of sequenced transcript reads identifies a particular cell type or cell sub-type; and

determining a transcription signature for each identified cell type or cell sub-type based at least in part on identifying differentially expressed transcripts between the identified cell type or cell sub-type by applying an unsupervised machine-learning model.

9. The method of claim 8, wherein the unsupervised machine learning model is a k-means model, a mixture model, a hierarchical clustering model, an anomaly detection model, a neural network model, an expectation-maximization (EM) model, a method of moments model, or a blind signal separation technique.

10. The method of claim 8, wherein the cell type and/or sub-types are retinal cell types and/or sub-types.

11. The method of claim 10, wherein the transcription signature identifies a rod bipolar cell, or one of the following cone bipolar cells;

BC1A, BC1B, BC2, BC3A, BC3B, BC4, BC5A, BCSB, BCSC, BCSD, (XBC), BC6, BC7, BC8, and BC9.

12. The method of claim 11, wherein the transcription signature identifies a bipolar cell and comprises at least one of Pcdh17, Pcdh10, Erbb4, Nnat, Col11a1, Sox6, Chrm2, Slitrk5, Lrrtm1, Cck, Lect1, Igfn1, Serpini1, Cpne9, Vstm2b, Casp7, Vsx1, Vsx2, Otx2, Scgn, Slc1a2, Slc6a9, Pax6, Tfap2a, Stx1b, Tacr3, Bhlhe23, Fat4, Fezf1, Fezf2, Wfdc2, My1k, Hs3st2, W1 s, Nxph1, Lhx3, Cabp5, Grm6, Kcng4, Cdh9, Htr3a, Hcn1, Hcn4, Nfia, Cntn5, BC046251, Gfra3, Areg, Kirrel3, Isl1, Spock3, Serpini1, Prkca, Car8, Sebox, Apoe, GluI, Aqp4, Cabp5, Syt2, Grik1, Prkar2b, Rcvrrn, Irx6, Gad1, Rho, Pdc, Nrl, Pde6a, Pde6h, Arr3, Opn1mw, and Opn1sw.