Methods and kits for identifying cancer treatment targets

1. A method comprising:

(a) introducing into a cancer cell and into a corresponding non-cancer cell of a same cell type as the cancer cell, a clustered regularly interspaced short palindromic repeat (CRISPR)/Cas system comprising:

i) a CRISPR/Cas effector polypeptide; and

ii) a CRISPR/Cas guide ribonucleic acid (RNA), or a nucleotide sequence encoding the CRISPR/Cas guide RNA, wherein the CRISPR/Cas guide RNA targets a target gene,

thereby producing a modified cancer cell and a modified non-cancer cell, both having reduced expression of the target gene;

(b) identifying, by a computer, a molecular feature or a phenotype in the modified cancer cell and the modified non-cancer cell, thereby generating data associated with the molecular feature or phenotype; and

(c) computer processing the data associated with the molecular feature or phenotype to determine that the molecular feature or phenotype is indicative of a reduction of cancerous state of the modified cancer cell,

wherein the determining comprises assessing an extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene,

wherein assessing the extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene comprises one or more of:

(i) using a geometric method to determine a distance between transcriptional profiles of the modified cancer cell and the modified non-cancer cell,

(ii) determining rank-order statistics for genes expressed in the modified cancer cell and the modified non-cancer cell,

(iii) processing gene expressions of the modified cancerous cell and the modified non-cancerous cell with a trained supervised machine learning classifier to classify the gene expressions as indicative of a cancer cell expression level or a non-cancer cell expression level, wherein the trained supervised machine learning classifier is trained on cancer cell expression data and non-cancer cell expression data, and

(iv) performing unsupervised clustering of transcriptional profiles of the modified cancerous cell and the modified non-cancerous cell.

2. The method of claim 1 , wherein the CRISPR/Cas effector polypeptide comprises a fusion protein comprising an enzymatically inactive CRISPR/Cas effector polypeptide and a transcriptional inhibitor.

3. The method of claim 1 , wherein the CRISPR/Cas effector polypeptide is enzymatically active, and wherein the CRISPR/Cas system comprises two CRISPR/Cas guide RNAs.

4. The method of claim 1 , wherein the non-cancer cell is a non-cancerous pancreatic cell, and wherein the cancer cell is a pancreatic cancer cell.

5. The method of claim 1 , wherein the non-cancer cell is a non-cancerous brain cell, and wherein the cancer cell is a brain cancer cell.

6. The method of claim 1 , wherein the non-cancer cell is a non-cancerous ovarian cell, and wherein the cancer cell is an ovarian cancer cell.

7. The method of claim 1 , wherein the non-cancer cell is a non-cancerous breast cell, and wherein the cancer cell is a breast cancer cell.

8. The method of claim 1 , wherein the molecular feature comprises an expression level of an indicator gene in the modified cancer cell and the modified non-cancer cell.

9. The method of claim 1 , wherein the identifying in (b) further comprises sequencing RNA obtained from the modified cancer cell and the modified non-cancer cell, thereby obtaining RNA sequence data therefrom.

10. The method of claim 9 , wherein the sequencing further comprises single-cell RNA sequencing.

11. The method of claim 9 , wherein (c) further comprises computer processing the RNA sequence data.

12. The method of claim 11 , wherein computer processing the RNA sequence data further comprises aligning the RNA sequence data to a reference genome.

13. The method of claim 1 , further comprising comparing the molecular feature or phenotype to a reference molecular feature or phenotype.

14. The method of claim 1 , further comprising computer processing the data associated with the molecular feature or phenotype using an expression extraction algorithm.

15. The method of claim 1 , wherein assessing the extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene further comprises using the geometric method to determine the distance between transcriptional profiles of the modified cancer cell and the modified non-cancer cell.

16. The method of claim 1 , wherein assessing the extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene further comprises processing gene expressions of the modified cancer cell and the modified non-cancer cell with the trained supervised machine learning classifier.

17. The method of claim 16 , wherein the trained supervised machine learning classifier comprises a deep learning classifier, a random forest classifier, a gradient boosted classifier, or a convolutional neural network.

18. The method of claim 17 , wherein the deep learning classifier further comprises a variational autoencoder.

19. The method of claim 17 , wherein the deep learning classifier further comprises a generative classifier.

20. The method of claim 19 , wherein the generative classifier comprises a non-linear mapping between (i) observed expression values of genes in the modified cancer cell and the modified non-cancer cell and (ii) a latent space.

21. The method of claim 20 , wherein the latent space is a low-dimensional latent space generated via non-linear feature reduction.

22. The method of claim 16 , further comprising performing a dimensionality reduction of the data associated with the molecular feature or phenotype.

23. The method of claim 1 , wherein the target gene comprises a plurality of target genes, and wherein (c) further comprises assessing an extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced combinatorial expression of the plurality of target genes.

24. The method of claim 1 , wherein the target gene is selected from the group consisting of: Ddit4, Cdkn2a, Hk2, Hes1, Asns, Galk1, Shmt2, Cct8, Gars, Psph, Ppid, Ruvbll, Chchd4, Nop16, Eif4ebpl, Gcsh, Ddx21, Ino80e, Tomm70a, Bri3 bp, Mpp6, Tomm20, Nhp2I1, Akr1b3, Noc2I, Nolc1, Tomm5, Nhp2, RsI24d1, Hnrnpdl, Dnajc2, Hacd1, Ddx3x, Matta, Ddx46, Gm16286, Tpi1, Gcat, Nmt1, Jun, Cbx3, Id3, Fam3c, Pcbp4, Id1, Mt2, Bcat1, Sparc, Pcolce, Ifitm3, S100a4, Xist, Tnfrsf26, Dusp9, Ly6a, Ccnd2, Emp3, Prkg2, Ndn, and a combination thereof.

25. The method of claim 1 , wherein the target gene is selected from the group consisting of: Gm10116, Pcbd1, Gamt, Gstm1, Chchd10, DIk1, Sod3, Bst1, Krt7, Anxa8, Slpi, Sorbs2, Ankrdl, Msln, KIra4, Igfbp7, Gm 10709, Tspan8, Gjb4, Anxa3, Krt19, Krt18, Akap12, Cdc42ep5, Tubb2a, FbIn2, Cyba, Timp3, Ucp2, Sgk1, Tubb2b, Fads3, H2-K1, Trp53, Spint2, Lsr, Prss2, Kcnk3, Vtn, Chga, Cpa1, Tm4sf4, Gc, Reg1, Try5, Ctrb1, Nts, Mdk, Bex2, Nkx6-2, Resp18, Cldn10, Penk, D930028M14R, Cela1, Rbp4, Bex4, Sepp1, Mest, Apoe, and a combination thereof.

26. The method of claim 1 , wherein assessing the extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene further comprises determining the rank-order statistics for genes expressed in the modified cancer cell and the modified non-cancer cell.

27. The method of claim 1 , wherein assessing the extent of transcriptional reversion toward a non-cancer cell expression level as a result of having reduced expression of the target gene further comprises performing the unsupervised clustering of the transcriptional profiles of the modified cancerous cell and the modified non-cancerous cell.

28. The method of claim 1 , further comprising identifying the target gene as a target for cancer treatment based at least in part on the determining in (c).