Alexander Pearlman, Ph.D.

For more than 15 years I have developed clinical-grade genomic diagnostic algorithms end to end — design, analytical validation, and regulatory readiness in oncology. My work centers on one question: whether tumor genomic copy-number alteration (CNA) patterns can predict metastatic potential across cancers.

At NYU Langone and then at Albert Einstein College of Medicine I developed a copy-number signature of metastatic potential in prostate cancer (Pearlman et al., 2012), extended it into a pan-cancer predictor of metastasis (Pearlman et al., 2018) and co-developed a next-generation sequencing assay for clinical specimens (Fofanov et al., 2019). This body of work is patented and published.

My validation philosophy is conservative: performance standards set against real-world outcomes rather than optimistic benchmarks, mandatory replication in independent cohorts, and a clear separation between what a test measures and what it has earned the right to claim. More recently, as an independent consultant, I have re-engineered the signature platform for the multimodal-data era, harmonizing genomic copy number alterations across whole-exome and whole-genome sequencing platforms. In preliminary, unpublished analyses of public data sets I have also harmonized copy number with RNA-seq gene expression from the same tumors and begun examining copy-number signal in ctDNA/cfDNA — an early step toward liquid-biopsy applications.

I maintain an active collaboration with the Department of Pathology at Albert Einstein College of Medicine, where I was an Instructor from 2011 to 2017; I was then an Associate in the Cancer Center and Department of Radiology (2017–2020) and a Consultant to the Cancer Center (2020–2022). I currently mentor bioinformatics students from Brown University and CUNY. From 2020 to 2023 I was Chief Scientific Officer of Hursana, LLC, where I led real-world-data programs, including a pharmacoepidemiology study of cannabis use and opioid prescribing in New York primary-care practices; Hursana holds no rights to the panMPS signature or assay.

1. Genomic copy-number signatures of metastatic potential, built for clinical specimens. I invented a clustering-based method for constructing genomic copy-number alteration (CNA) signatures and used it to build a prognostic signature that predicts the metastatic potential of prostate cancer from clinical specimens, including needle biopsies (Pearlman et al., 2012). I subsequently led the extension of this approach into a robust pan-cancer copy-number predictor of metastasis, replicated in independent retrospective cohorts across multiple epithelial cancers (Pearlman et al., 2018), with mechanistic follow-up identifying ectopic Otoconin 90 expression associated with metastasis functions in triple-negative breast cancer (Pearlman et al., 2019). To make the signature work on the small, imperfect samples real patients provide, I co-developed a rapid next-generation sequencing method for prediction of prostate cancer risks designed for clinical workflows (Fofanov et al., 2019). The signature and assay are patented (US 10,519,505; US 12,227,809). As an independent consultant I continue to develop this platform for the multimodal-data era.
2. Functional assays for inherited cancer risk. With colleagues at Albert Einstein College of Medicine, I contributed to the development of functional variant analyses (FVAs) — flow-cytometric functional readouts of DNA-repair pathway variants — demonstrating that they predict pathogenicity in the BRCA1 DNA double-strand-break repair pathway (Loke et al., 2015) and that flow-variant analysis of circulating peripheral-blood B cells predicts breast cancer risk (Syeda et al., 2017). This work underlies a patented functional assay for assessing germline cancer risk from DNA-repair variants (US 12,253,524), on which I am a co-inventor.
3. Population genetics and genetic ancestry. Earlier in my career I contributed to foundational studies of Jewish population structure, including the demonstration that major Jewish diaspora populations comprise distinct genetic clusters with shared Middle Eastern ancestry (Atzmon et al., 2010) and that North African Jewish and non-Jewish populations form distinctive, orthogonal clusters (Campbell et al., 2012). This work informs ancestry-aware genomic analysis — an important consideration for building cancer predictors that generalize across populations.
4. Metabolic interventions in cancer. As Co-Principal Investigator of a translational presurgical ketogenic-diet breast-cancer study at Albert Einstein College of Medicine, I coordinated clinical, pathology, and tumor-omics data across the study's human and mouse arms, serving as the cross-functional bridge among scientific, clinical, and analytical groups. The mouse-arm work demonstrated the effect of a ketogenic diet and its synergy with rapamycin in a model of breast cancer (Zou et al., 2020).