Search

EP-4741516-A1 - METHODS AND SYSTEMS FOR CLASSIFYING CANCER AND DETECTING IMPROVED CANCER THERAPIES

EP4741516A1EP 4741516 A1EP4741516 A1EP 4741516A1EP-4741516-A1

Abstract

Disclosed herein are methods (100) and systems (200) for classifying a cancer from a subject. The methods (100) and systems (200) classify the cancer based on similar characteristics, e.g., molecular profiles. The methods (100) and systems (200) may be predictive of the subject's response to treatments based on the classification of the cancer. The methods (100) and systems (200) may be used to define improved therapies for subjects with cancers with limited treatment options, e.g., rare cancers.

Inventors

  • CORCORAN, Emma Tung
  • SELITSKY, Sara

Assignees

  • Tempus AI, Inc.

Dates

Publication Date
20260513
Application Date
20251111

Claims (15)

  1. A method of classifying a cancer from a subject: obtaining, with a computer system, sequencing read data collected from a sample from the cancer of the subject, the read data comprising RNA sequencing data; classifying, with the computer system, the cancer as a subtype of cancer, using a trained machine learning algorithm, wherein the subtype of cancer comprises a plurality of cell proliferative diseases with common characteristics, wherein the common characteristics comprise similar molecular profiles, wherein the trained machine learning algorithm is trained on a data set of sequencing read data collected from a cohort of subjects suffering from cancer, wherein the squamous cell carcinomas comprises anogenital, cervical, esophageal, head and neck, lung, skin, urothelial, colorectal, and vulvar squamous cell carcinomas.
  2. The method of claim 1, wherein the sample comprises at least one of a tumor sample, blood sample, or cell free DNA.
  3. The method of claim 1 or claim 2, wherein the plurality of cell proliferative diseases comprises squamous cell carcinomas (SCC).
  4. The method of any preceding claim, wherein the common characteristics further comprises similar phenotypes, prognosis, and predicted responses to treatment; and optionally wherein the predicted response to treatment comprises predicted response to chemotherapy.
  5. The method of any preceding claim, wherein the similar molecular profiles comprise expression levels of one or more of RNF186, CCL15, TMIGD1, RPL10L, ATOH1, ANKS4B, ALPI, SCL17A4, B3GNT6, MOGAT3, SFTA3, GGTLCl, NAPSA, SFTPD, MS4A15, VWA3A, ANKRD66, HABP2, CPAMD8, KCNK3, CFAP95, CFAP43, OSGIN1, SRXN1, G6PD, ETNK2, DGKG, NDGA1, LDC1, RAB3B, TAGA3, PLCXD2, GSTM2, WNT5A, RAB25, TTLL10, SGPP2, SPINK9, IGSF9, ARHGEF26, PIR, RAPGEFL1, CIMAP2, SCNN1A, ZBTB7C, BDNF, ARG1, TREX2, CMA1, KRTAP5-4, LIPM, SPTLC3, GCSAML, HAL, LGALSL, VSIG8, TMC4, ELMOD1, SMPD3, GRACDL, DPF1, RAX, GATM, KLHL35, TMEM236, ACTBL2, TCEA3, EPB41LB, CT62, DKK3, FJX1, CASP5, MANEAL, or NUP210.
  6. The method of any preceding claim, wherein the cohort of subjects comprises subjects diagnosed with at least 5 different types of cancers.
  7. The method of any preceding claim, wherein each subject in the cohort of subjects has been diagnosed with a squamous cell carcinoma.
  8. The method of any preceding claim, wherein the trained machine learning algorithm comprises at least one of a gradient boosting model, a random forest model, a neural network, a regression model, ElasticNet, or a Naive Bayes model; and optionally wherein the trained machine learning algorithm is ElasticNet.
  9. The method of any preceding claim, wherein the method further comprises generating a report; and optionally wherein the report comprises the subtype of cancer, the plurality of cell proliferative diseases with common characteristics, and the molecular profiles.
  10. The method of any preceding claim, wherein the report further comprises a list of treatment options.
  11. The method of any preceding claim, wherein the cancer is classified as a squamous cell carcinoma; and/or the method of any of preceding directly or indirectly dependent upon claim 3, wherein the cancer is not classified as a squamous cell carcinoma.
  12. The method of any preceding claim, wherein the treatment options are identified based on the plurality of cell proliferative diseases with common characteristics and the molecular profile.
  13. The method of any preceding claim, wherein the cancer has limited treatments comprising at least one of ineffective treatments, few treatments, and no known treatments; and optionally wherein the cancer with limited treatments is vulvar squamous cell carcinoma.
  14. The method of any preceding claim, wherein the plurality of signature genes comprise two or more genes selected from one of (i), (ii), (iii), (iv), (v), or (vi): (i) CRACDL, DPF1, RAX, GATM, KLHL35, TMEM236, ACTBL2, TCEA3, EPB41L4B, CT62, DKK3, FJX1, CASP5, MANEAL, NUP210, RPL10L, FOXF2, LIPG, GRID2, C2orf48, SH3TC2, MECOM, SPACA5, SHC4, R3HDML, BRME1, L1TD1, ZAR1, SLC28A1, FAM169A, FEV, SPMIP11, GLI1, CRYBB2, KIRREL3, PI15, FEZ1, C2CD4B, PLEKHG4, GOLGA6L10, GRIN2C, CELF5, TSPAN18, CARD10, ACOD1, PLCH1, AR, MTNR1A, PPP1R14C, B4GALNT3, ESR1, PITX1, PRSS46P, CHRNA3, DNAJB13, RET, PAX8, ANKRD65, ZDHHC19, IGF2BP2, KLF8, TACSTD2, CCDC166, TRIL, ZP4, SHISAL2A, TMT1B, ADGRE1, OCM, PIWIL2, SNCB, PDPN, RASD2, NICOL1, COLEC10, GJE1, EGR3, RIBC2, SLC26A5, SLC2A12, GABRB1, SGCG, GABRA2, FAM81A, ATP8A2, USP2, RAPGEFL1, NAALADL2, CCDC185, NANOG, HTR2C, SLC10A4, PHACTR3, NPSR1, TRH, PMP2, HBEGF, C22orf31, LVRN, or ZSWIM5; (ii) ARG1, TREX2, CMA1, KRTAP5-4, LIPM, SPTLC3, GCSAML, HAL, LGALSL, VSIG8, TMC4, ELMOD1, SMPD3, ACER1, ABCG4, ATP6V1C2, TPPP2, DCD, ELOVL4, KRT25, RNF222, ACSBG1, ANKRD31, MELTF, NPM2, FRMPD1, ENDOU, LCE5A, USP2, LCE1B, DGAT2, LCE1E, PNPLA1, SERPINA12, SYT17, TMEM45A, CCL27, LCE6A, RDH12, ASPRV1, XKRX, TUBB2A, MMP27, HOPX, MS4A2, KRT33B, ESYT3, GALNT6, DEGS2, LIPN, IL37, ACKR2, LCE1D, HTR3A, DCT, RARB, OPN1MW, SPAG11B, FLG2, DEFB105B, VIPR1, LCE1A, SPACA5, SCGB1D2, GLB1L3, TEX28P2, HDC, PTGS1, RDH16, KRT80, CIDEA, SCN4B, HYAL4, CTSG, GPR63, TYR, LELP1, LYPD5, SCGB2A2, HOXD1, TEX28P1, RHBG, FLG, AADACL3, BPIFC, TRPM1, OPN1LW, NEU2, NSG1, MECOM, GALNT12, COX8C, TEX28, IL1F10, LORICRIN, GATA3, PTPN5, NWD2, KRT84, or WNT16; (iii) RAB25, TTLL10, SGPP2, SPINK9, IGSF9, ARHGEF26, PIR, RAPGEFL1, CIMAP2, SCNN1A, ZBTB7C, BDNF, ACSBG1, PGAP4, ZNF711, ACP3, TMEM125, CLDN4, GGT6, P2RY1, Clorf210, OTX1, CSN3, ESYT3, TTC39A, RNF183, VSIG8, DNAI7, C22orf31, FAM181A, GSTA4, ALG1L2, PLS1, BMP7, CFAP73, EFCC1, ISL2, ENDOU, L1CAM, CYP4X1, GPX2, IL20RA, COMMD5P1, SOX1, PCP4L1, KRTAP5-2, FA2H, SAMD12, SRXN1, GRID2, TRH, TLCD4-RWDD3, RNF225, MCIDAS, NDRG4, PRR35, CCN3, LIPM, OVOL2, CGN, POU2F3, HOPX, DOC2B, RBBP8NL, B4GALNT3, SPOCK1, GLYATL1, SRRM3, BSPRY, CACNA2D3, PHGDH, BCL2L15, B3GNT6, ZNF385C, VEGFC, EBF3, ACTBL2, VAX2, ZDHHC11, ART3, MYH14, TGFBI, C2orf48, LINC02898, CFAP276, PLA2G3, GCSAML, MYOM3, FGFR2, ALG1L1P, KLHDC7A, OPRKl, POF1B, CBX2, CEACAM1, THBS1, NEBL, CCDC185, C20orf144, or CHODL; (iv) OSGIN1, SRXN1, G6PD, ETNK2, DGKG, MDGA1, ODC1, RAB3B, GATA3, PLCXD2, GSTM2, WNT5A, BDNF, PIR, OR6C2, ME1, GPAT3, NQOl, TRIM16L, JAKMIP3, NECAB2, GLI2, SLC38A8, CYP2S1, GSTM3, CCL28, GPX2, NOG, C1QTNF12, TSPAN7, OR56B4, SCN9A, NKX6-1, GLI1, PANX2, CFAP20DC, Clorf226, ENTHD1, SLC7A11, UGT1A1, MST1R, AKRIC1, RAB6B, H4C9, CCDC125, VPS37D, DPF1, SLC6A13, B4GALNT3, GCNT2, GASK1A, CCL26, NR0B1, KLRG1, ARTN, NRCAM, ELAPOR2, KCND3, TPRG1, ZMAT1, OTOP2, RORC, PCYT1B, RND2, SGCZ, SAMD12, HAP1, BRD2, DAZ3, AKR1C3, ENPP3, ANO1, MACROD2, UPK1B, JAKMIP2, AKR1C4, ETNPPL, PFN2, ANXA10, LRRC2, ZDHHC2, NUDT11, CNTN6, SLC4A3, ALDH3A1, TMC1, OR6C70, DLG2, CIMAP2, VIPR1, SPTLC3, KIT, CYP26A1, ROR1, PMP2, NYAP1, FGF13, SAMD3, S100A5, or LGSN; (v) SFTA3, GGTLCl, NAPSA, SFTPD, MS4A15, VWA3A, ANKRD66, HABP2, CPAMD8, KCNK3, CFAP95, CFAP43, CFAP221, NKX2-1, FOXB1, C16orf89, C8B, NEK5, LRP2, AQP4, SLC9C2, C4BPA, TMEM212, STOML3, CDH7, KIAA2012, DLG2, TTC29, USP44, F11, PPM1H, PGC, SFTPB, ODAD1, CATSPERD, PEBP4, PLCH1, ZBBX, CFAP107, C1orf87, DAW1, ROPN1L, FYB2, KCTD16, C8orf34, PCDHAC2, CP, ERICH3, RP1, ABCC6, KHDRBS2, PLA2G1B, SPEF2, SCN1A, CFAP276, WFDC6, SLC22A31, RGPD3, KRTAP10-9, DNAI1, ACSM1, RAB6C, CFAP65, MARCHF10, CDHR3, FRMPD2, DNAI7, ERICH2, DNAH12, ZNF648, CIMIP1, GARIN6, ARMC3, HOATZ, C2orf73, C1orf222, TEKT2, CFAP90, AGBL1, SNTN, DRC1, MIA2, C4A, RSPHI1, ASB4, STMND1, DNAH5, CABCOCO1, NME5, HP, TSPAN19, CGNL1, MALRD1, SHISA3, CNTN6, SCGB3A2, NRGN, XAGE1C, ABCA3, or HYDIN; (vi) RNF186, CCL15, TMIGD1, RPL10L, ATOH1, ANKS4B, ALPI, SLC17A4, B3GNT6, MOGAT3, NR1I2, IHH, MS4A12, A1CF, FEV, CLRN3, NHERF4, INSL5, R3HDML, GUCA2B, NXPE1, MYO1A, HNF1A, NAT2, PYY, NXPE4, AQP8, NOX1, REG3A, UGT2A3, TRIM15, B3GALT1, ISX, CDH17, NXPE2, MEP1A, GCG, CDHR2, CHST5, B3GNT7, ZG16, GALNT8, EFNA2, TINAG, LYPD8, SLC51B, FABP2, LEFTY1, HTR4, CHGA, TM4SF5, MYO7B, LGALS4, SLC6A19, CDX1, SI, RETNLB, PLA2G10, BCL2L15, TMEM236, SLC18A1, SAMD13, CA7, HHLA2, SULT1B1, C5orf52, GPA33, REG1B, GP9, HEPACAM2, LRRC31, GUCA2A, REG4, VSIG2, CLCAl, SLC26A3, IYD, BNIP5, GREM2, SGK2, HGD, VIL1, VSTM2A, KRT20, SPMIP10, SLC28A2, AOC1, ANXA13, GUCY2C, FAM135B, CA1, CAPN9, GABRA2, ALDOB, SULT1C3, HNF4A, MUC12, PPP1R14D, SPINK4, or BTNL3.
  15. A system for classifying a cancer from a subject, the system comprising at least one memory, and at least one processor coupled to the at least one memory, the system configured to cause the at least one processor to execute instructions stored in the at least one memory to: obtain, with a computer system, sequencing read data collected from a sample from the cancer of the subject, the read data comprising RNA sequencing data; classify, with the computer system, the cancer as a subtype of cancer, using a trained machine learning algorithm, wherein the subtype of cancer comprises a plurality of cell proliferative diseases with common characteristics, wherein the common characteristics comprise similar molecular profiles, wherein the trained machine learning algorithm is trained on a data set of sequencing read data collected from a cohort of subjects suffering from cancer.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS The present application claims priority to U.S. Provisional Patent Application No. 63/719,617 that was filed November 12, 2024, the entire contents of which are hereby incorporated by reference. TECHNICAL FIELD This present disclosure relates to systems, methods, and compositions useful for profiling a subject's cancer by classifying the cancer by a particular cancer subtype. The present disclosure also relates to systems and methods for diagnosing, matching a patient with appropriate treatments, monitoring, or predicting disease, condition, or therapeutic outcomes based on the cancer subtype of a subject. BACKGROUND Squamous cell carcinomas (SCCs) can occur in a variety of tissues with varying frequencies. Rare cancers are unlikely to be the subject of clinical trials, in part, due to the difficulty of recruiting a sufficient subject population. The limited number of clinical trials further complicates the diagnosis and treatment of these diseases, SCCs in different tissue types may have similar morphologies. Therefore, there is a need in the art for methods to characterize SCCs, and other cancers, based on their molecular profile which may lead to improved diagnostics, improved treatment options, and improved recruiting of subjects with rare cancers into clinical trials. SUMMARY To the accomplishment of the foregoing and related ends, the invention, then, comprises the features hereinafter fully described. The following description and the annexed drawings set forth in detail certain illustrative aspects of the invention. However, these aspects are indicative of but a few of the various ways in which the principles of the invention can be employed. Other aspects, advantages and novel features of the invention will become apparent from the following detailed description of the invention when considered in conjunction with the drawings. In an aspect of the current disclosure, methods are provided. In some embodiments, the methods comprise: obtaining, with a computer system, sequencing read data collected from a sample from the cancer of the subject, the read data comprising RNA sequencing data; classifying, with the computer system, the cancer as a subtype of cancer, using a trained machine learning algorithm, wherein the subtype of cancer comprises a plurality of cell proliferative diseases with common characteristics, wherein the common characteristics comprise similar molecular profiles, wherein the trained machine learning algorithm is trained on a data set of sequencing read data collected from a cohort of subjects suffering from cancer. In some embodiments, the sample comprises at least one of a tumor sample, blood sample, or cell free DNA. In some embodiments, the plurality of cell proliferative diseases includes squamous cell carcinomas (SCC). In some embodiments, the SCC includes anogenital, cervical, esophageal, head and neck, lung, skin, urothelial, colorectal, and vulvar squamous cell carcinomas. In some embodiments the common characteristics further include similar phenotypes, prognosis, and predicted responses to treatment. In some embodiments, the similar molecular profiles comprise expression levels of one or more of RNF186, CCL15, TMIGD1, RPL10L, ATOH1, ANKS4B, ALPI, SCL17A4, B3GNT6, MOGAT3, SFTA3, GGTLC1, NAPSA, SFTPD, MS4A15, VWA3A, ANKRD66, HABP2, CPAMD8, KCNK3, CFAP95, CFAP43, OSGIN1, SRXN1, G6PD, ETNK2, DGKG, NDGA1, LDC1, RAB3B, TAGA3, PLCXD2, GSTM2, WNT5A, RAB25, TTLL10, SGPP2, SPINK9, IGSF9, ARHGEF26, PIR, RAPGEFL1, CIMAP2, SCNN1A, ZBTB7C, BDNF, ARG1, TREX2, CMA1, KRTAP5-4, LIPM, SPTLC3, GCSAML, HAL, LGALSL, VSIG8, TMC4, ELMOD1, SMPD3, GRACDL, DPF1, RAX, GATM, KLHL35, TMEM236, ACTBL2, TCEA3, EPB41LB, CT62, DKK3, FJX1, CASP5, MANEAL, or NUP210. In some embodiments, the cohort of subjects comprises subjects diagnosed with at least 5 different types of cancers. In some embodiments, each subject in the cohort of subjects has been diagnosed with a squamous cell carcinoma. In some embodiments, the trained machine learning algorithm comprises at least one of a gradient boosting model, a random forest model, a neural network, a regression model, ElasticNet, or a Naive Bayes model. In some embodiments, the method further comprises generating a report. The report may include the subtype of cancer, the plurality of cell proliferative diseases with common characteristics, and the molecular profiles. The report may further include a list of treatment options. In some embodiments, treatment options are identified based on the plurality of cell proliferative diseases with common characteristics and the molecular profiles. In some embodiments the cancer may have limited treatment options comprising at least one of ineffective treatments, few treatments, and no known treatments. In some embodiments the cancer with little limited treatments is vulvar squamous cell carcinoma. In some embodiments, the molecular profiles comprise RNA expression data and the compute