ORDER A TEST Log In Contact Us

LIFE SCIENCES /// ALGORITHMIC TESTS

Powering smarter, faster drug development with AI-driven algorithms

Harness multimodal data, AI/machine learning (ML), and bioinformatics expertise to help discover, validate, and deploy molecular algorithms.

Precision medicine is evolving

The future of precision medicine relies on more than traditional biomarkers—it requires a new approach. AI-driven algorithms that leverage real-world data (RWD) are driving the next generation of drug development by helping:

  • Predict response to therapies with greater accuracy
  • Uncover novel biomarkers to drive targeted therapy development
  • Identify patients for clinical trials based on molecular and clinical criteria
  • Inform companion diagnostic (CDx) development to support regulatory and clinical adoption
  • Optimize treatment selection by integrating real-world and multimodal data

Our systematic approach to developing molecular algorithms

  1. Data-driven discovery

    Tempus ML experts analyze deep multimodal datasets—including 8M+ de-identified research records, 1M+ matched clinical-genomic records, and 2M+ imaging records—to detect emerging patterns in disease biology, treatment response, and patient stratification.

  2. AI-powered refinement

    AI/ML models validate and refine biomarker hypotheses by evaluating co-expression patterns, resistance mechanisms, and clinical factors, and assess clinical impact to support biomarker-driven clinical trials and targeted therapy development.

  3. Real-world validation

    Our team uses real-world patient cohorts to evaluate and validate biomarker performance and clinical utility, helping life science partners generate the evidence needed for regulatory submissions and clinical adoption.

  4. Seamless integration into trials & CDx development

    Once validated, biomarkers can be incorporated into clinical trials, refined for CDx development with support from our regulatory expertise, and deployed to practicing clinicians at scale by leveraging our network of 6,500+ oncologists.

CASE STUDY

From insight to impact: The development of IPS

Tempus leveraged DNA, RNA, and de-identified clinical data from clinical testing to develop, validate, and launch the Immune Profile Score (IPS) algorithm as a laboratory-developed test in ~18 months.

Learn more about IPS
  • Uncovering unmet need While PD-L1, TMB, and MSS are commonly used biomarkers for guiding immunotherapy, they have limitations, including suboptimal patient stratification and inconsistent response prediction.
  • Leveraging RWD To address these gaps, Tempus analyzed 100+ immune-related DNA and RNA features, including inflammatory biomarkers, immune resistance mechanisms, and tumor proliferation signals, from over 1,700 patients treated with immune checkpoint inhibitors (ICIs).
  • Applying AI to unlock deeper insights Advanced ML models were applied across DNA, RNA, and immunohistochemistry (IHC) data to uncover predictive patterns, refining IPS into a powerful prognostic tool for ICI response.
  • Real-world validation & clinical impact IPS demonstrated strong prognostic utility in Tempus’ large, de-identified retrospective pan-cancer cohort, with IPS-High patients showing significantly longer overall survival compared to IPS-Low patients across multiple subgroups, including TMB-low and MSS patients.1
  • IPS Deployment IPS can be deployed in a research setting to increase confidence in identifying potential responders, improve patient stratification, and optimize study designs. Additionally, IPS has the potential to be developed as a CDx for existing or novel therapies.

Tempus’ algorithmic tests for life science partners

  • IPS

    A DNA and RNA-based molecular signature that provides prognostic insights into real-world overall survival (rwOS) outcomes following ICI treatment in metastatic solid tumors.

    May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination.

    Both DNA and xR RNA seq are required for processing. No additional tissue required.

    Learn more
    Download validation summary
  • PurISTTM

    Identifies the molecular subtype of patients with unresectable stage III or stage IV pancreatic ductal adenocarcinoma (PDAC).

    • Utilizes RNA sequencing information to classify patients with PDAC into either a basal or classical subtype.
    • Patients with a classical subtype were found to have an overall better prognosis and superior median overall survival when treated with first-line FOLFIRINOX, compared to patients with a basal subtype.2,3

    May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

    xR RNA seq is required for processing. No additional tissue required.

    Download validation summary
  • CMS

    Supports molecular stratification of colorectal cancer (CRC) patients by assigning Consensus Molecular Subtypes (CMS).

    Enables classification of both primary and metastatic tumors to enhance patient selection and stratification in CRC studies.4

    May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

    xR RNA seq is required for processing. No additional tissue required.

  • HRD

    Predicts the probability of a patient’s cancer having a phenotype characterized by the inability to repair DNA breaks via the homologous recombination repair (HRR) pathway, known as homologous recombination deficiency (HRD).5

    • For ovarian and breast cancer, where DNA-based methods of HRD detection are common or under investigation, Tempus HRD provides a result based on DNA genome-wide loss of heterozygosity (GWLOH) or evidence of biallelic BRCA1 or BRCA2 loss from the xT CDx test.
    • For patients with other cancers, Tempus HRD provides an HRD score based on whole transcriptome RNA expression using data from the xR test.

    May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

    The matched normal sample for xT CDx is required for breast and ovarian cancer. xR RNA seq is required for processing in all other cancers. No additional tissue required.

    Download validation summary
  • TUMOR ORIGIN (TO)

    Uses tumor RNA expression results to predict the patient’s most likely cancer type(s) from 68 possible diagnoses.

    • Intended for patients with cancer of unknown primary (CUP).

    May be ordered as an add-on with xT CDx & xR combination OR xT & xR Combination OR xR RNA seq.

    xR RNA seq is required for processing. No additional tissue required.

    Download validation summary

A DNA and RNA-based molecular signature that provides prognostic insights into real-world overall survival (rwOS) outcomes following ICI treatment in metastatic solid tumors.

May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination.

Both DNA and xR RNA seq are required for processing. No additional tissue required.

Learn more
Download validation summary

Identifies the molecular subtype of patients with unresectable stage III or stage IV pancreatic ductal adenocarcinoma (PDAC).

  • Utilizes RNA sequencing information to classify patients with PDAC into either a basal or classical subtype.
  • Patients with a classical subtype were found to have an overall better prognosis and superior median overall survival when treated with first-line FOLFIRINOX, compared to patients with a basal subtype.2,3

May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

xR RNA seq is required for processing. No additional tissue required.

Download validation summary

Supports molecular stratification of colorectal cancer (CRC) patients by assigning Consensus Molecular Subtypes (CMS).

Enables classification of both primary and metastatic tumors to enhance patient selection and stratification in CRC studies.4

May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

xR RNA seq is required for processing. No additional tissue required.

Predicts the probability of a patient’s cancer having a phenotype characterized by the inability to repair DNA breaks via the homologous recombination repair (HRR) pathway, known as homologous recombination deficiency (HRD).5

  • For ovarian and breast cancer, where DNA-based methods of HRD detection are common or under investigation, Tempus HRD provides a result based on DNA genome-wide loss of heterozygosity (GWLOH) or evidence of biallelic BRCA1 or BRCA2 loss from the xT CDx test.
  • For patients with other cancers, Tempus HRD provides an HRD score based on whole transcriptome RNA expression using data from the xR test.

May be ordered as an add-on with xT CDx & xR combination OR xT & xR combination OR xR RNA seq.

The matched normal sample for xT CDx is required for breast and ovarian cancer. xR RNA seq is required for processing in all other cancers. No additional tissue required.

Download validation summary

Uses tumor RNA expression results to predict the patient’s most likely cancer type(s) from 68 possible diagnoses.

  • Intended for patients with cancer of unknown primary (CUP).

May be ordered as an add-on with xT CDx & xR combination OR xT & xR Combination OR xR RNA seq.

xR RNA seq is required for processing. No additional tissue required.

Download validation summary
  1. Zander AD, Erbe R, Liu Y, et al. Clinical validation of a novel multi-omic algorithm for stratifying outcomes in a real-world cohort of metastatic solid cancer patients treated with immune checkpoint inhibitors. J Immunother Cancer. 2024;12:doi: 10.1136/jitc-2024-SITC2024.0188.
  2. Rashid NU, et al. Purity independent subtyping of tumors (PurIST), a clinically robust, single-sample classifier for tumor subtyping in pancreatic cancer. Clin Cancer Res. 2020;26(1):82-92. doi:10.1158/1078-0432.CCR-19-1467.
  3. Wenric S, Davison JM, Guittar J, et al. Real-world data validation of the purist pancreatic ductal adenocarcinoma gene expression classifier and its prognostic implications. medRxiv. Published online February 24, 2023. doi:10.1101/2023.02.23.23286356.
  4. Sedgewick AJ, Chandra T, Taxter T, Guinney J. Abstract 5059: Robust single sample consensus molecular subtype classification for primary and metastatic colorectal cancer. Cancer Res. 2024;84(6_suppl):5059. doi:10.1158/1538-7445.AM2024-5059.
  5. Leibowitz BD, Dougherty BV, Bell JSK, et al. Validation of genomic and transcriptomic models of homologous recombination deficiency in a real-world pan-cancer cohort. BMC Cancer. 2022;22(1):587. doi:10.1186/s12885-022-09669-z.

Partner with us

Our expertise is in transforming raw data into real-world impact. Let’s build the next breakthrough together.