NEW YORK (GenomeWeb) – Precision for Medicine has launched the PATH analytics platform, a new product for the translational research market that's designed to help users combine and use clinical and genomic information to stratify patients as part of efforts to develop more tailored therapies.
Precision for Medicine provides a wide range of tools and services that support research and clinical development of targeted therapies, diagnostics, clinical solutions, biomarker analytics, and regulatory guidance.
PATH, one of the company's newest offerings, is a multi-model analytics platform that offers tools for integrating and visualizing information for biomarker discovery and validation, identifying patient-specific biomarker signatures, and stratifying patients. Users can interact with data in real time, integrate public/proprietary datasets, model and explore hypothetical scenarios, and more.
The company unveiled the platform in February at the Cambridge Healthtech Institute Molecular Medicine Tri-Conference in San Francisco. It plans to formally launch the software sometime this quarter. However the system is live and the company is accepting customers, Scott Marshall, one of the managing directors for Precision for Medicine's analytics practice, told GenomeWeb.
In fact, the company has already begun working with some unnamed large pharmaceutical companies who are using PATH in their personalized medicine projects. However, the company also believes that the tool would be of benefit to molecular diagnostics companies and plans to target those as well, Marshall said.
According to Marshall, his firm launched the platform to help translational researchers better handle and use the large quantities of complex data that are generated by high-throughput assays to, for example, stratify patients in clinical trials.
"What we see is that almost daily there are new technologies coming out that are developing and generating high-throughput biomarker data and the scale of that is continuing to increase and the cost to decrease," he said.
"At the same time, what we've seen ... is that maybe the analytics is not advancing at the same pace as the technology in terms of data generation," he added. "What we wanted to do is build from the ground up an analytics platform that is specifically designed to address the complexities that come with high-throughput biomarker assays and simultaneously address the objectives of any personalized medicine programs."
According to Marshall, the system is separated into two modalities. The first modality applies the company's proprietary advanced analytics algorithms to input data, and the second modality puts the output of the first into interactive tables and visuals that enable real-time data exploration, he said.
PATH accepts biomarker data from next-generation sequencing assays, microarrays, and multiplex immunoassays; clinical information such as patient treatment and outcomes data; and demographic information like age and gender.
The system runs initial preprocessing and quality control steps on the inputted data, as well as an initial filtering step for high-dimensional data. The tool then applies Bayesian hierarchical modeling and latent variable estimation techniques to input data and does things like measure individual biomarker contributions and estimate the magnitude of those contributions for biomarker discovery. These methods are also used to characterize uncertainty around patient-specific biomarker signatures; estimate uncertainty around clinical and diagnostic performance metrics for patient stratification purposes; and more.
Specifically, the analytics framework takes biological data such as somatic mutations and RNA-seq data and then combines it into a single modeling framework, Marshall explained. Next, it selects the biomarkers most relevant for impacting treatment response and then estimates a patient-specific molecular signature based on those markers — a so-called PATH index — and uses that to classify patients based on how they would respond to a given treatment, for instance.
Information from the first modality feeds into the second. In this modality, the list of biomarkers gleaned from the analysis in modality one are stored in searchable and sortable web-linked tables and represented in graphs and figures that can be explored in real time. The tables are linked to genome browsers and public annotation databases, such as those maintained by the National Center for Biotechnology Information and dbSNP, for additional contextual information such as relevant gene pathways or gene function information. They can also be linked to private customer-created databases. The interface to this modality can be customized to fit specific clients' needs and the resulting tables and visuals are easily exported for use in presentations and documents.
Precision for Medicine currently offers web-based access only to PATH’s second modality. The company is also developing a PATH mobile option which will run on select tablet and mobile devices. It expects to release PATH mobile in the fourth quarter of this year. It also plans to integrate additional annotation bases and genome browsers with the system, Marshall said.
The company declined to disclose details about its business model and pricing for PATH.