DEVELOPMENT OF A LUNG-CANCER SUBTYPING DIAGNOSTIC Abstract The overall goal of the proposed project will be to create a high sensitivity, high specificity subtyping test for lung cancer. Lung cancer subtypes differ significantly in pathology, metastatic potential, therapy selection and response. Lung cancer subtyping is critical for selecting the most effective treatment and thus reducing overall mortality. This subtyping capability will be added to Orbit Genomics’ first lung cancer diagnostic test that is an aid-to-diagnosis test for positive LDCTs with indeterminate pulmonary nodules. It distinguishes benign from malignant pulmonary nodules with a prototype efficacy of 92% Positive Predictive Value (PPV) and 97% Negative Predictive Value (NPV). Subtyping capability will markedly improve the value of the existing test which is in validation trials now using samples from Drs. Yankelevitz and Henschke at the Icahn School of Medicine at Mt. Sinai and is on a course to be launched as a Lab Developed Test (LDT) next year. Our first diagnostic test will supplement LDCT imaging to achieve unprecedented overall diagnostic accuracies and should prevent 500,000 unnecessary biopsies in the US annually, saving $6 billion in procedural costs. The lung cancer subtyping diagnostic test will expand the use of the test to all pulmonary nodules while providing additional information needed for treatment, further reducing the need for biopsies and improving patient outcomes while saving money. Clinical samples from Mt. Sinai with pathology determined lung cancer subtypes were used to identify statistically significant subtype informative loci and subtype classification in preliminary experiments. In previous publications we showed that an earlier version of our approach could provide brain cancer subtyping information. In order to bring this novel lung cancer subtyping diagnostic test to the market we will need to obtain subtype training samples (n=700, patient blood samples) from our existing and expanded network of clinical collaborators (SA1); to upgrade microsatellite allelotyper/genotypers, our informative loci selection and data mapping-to-image-space algorithms and train our new, novel AI Image-based classifier on the exome sequences of samples with known subtypes (small cell lung carcinoma (SCLC), non- small cell lung carcinoma (NSCLC) and its further subtypes (Squamous cell carcinoma; Adenocarcinoma; Large cell carcinoma)) (SA2). While creating this new version of the test we will iteratively optimize SA1-2 performance and then validate it (SA3) on independently obtained, pathologically well-characterized, clinically diagnosed patient blood samples (n=200) analyzed in Orbit Genomics’ CLIA laboratory. Such a subtype capable diagnostic test will be launched as an LDT, and will provide fast, accurate results to treating clinicians, ultimately saving lives, patient suffering and money.