Computational analysis of biological data using manifold and a hyperplane

Inventors

Eden, Eran • Oved, Kfir • NAVON, Roy • COHEN-DOTAN, Assaf • BOICO, Olga

Assignees

Memed Diagnostics Ltd

Abstract

A method of analyzing biological data containing expression values of a plurality of polypeptides in the blood of a subject. The method comprises: calculating a distance between a segment of a curved line and an axis defined by a direction, the distance being calculated at a point over the curved line defined by a coordinate along the direction. The method further comprises correlating the distance to the presence of, absence of, or likelihood that the subject has, a bacterial infection. The coordinate is defined by a combination of the expression values, wherein at least 90% of the segment is between a lower bound line and an upper bound line.

Core Innovation

The patent describes a computational framework for analyzing biological data by using expression levels of blood polypeptides as biomarkers. It receives biomarker values in the blood of a subject and calculates a score responsively to a user activation of a calculation activation control, then maps the score to a graphical likelihood presentation on a graphical user interface (GUI).

The disclosed invention models biomarker-derived data as a manifold and computes a perpendicular distance from a curved segment or curved surface to a hyperplane described by an axis or plane. Coordinates on the curved segment or surface are determined using optionally nonlinear combinations of biomarker expression values and associated direction vectors, and probabilistic classification functions for likelihoods are correlated with the computed distance and with bounding rules using lower and upper bound curves or surfaces.

Likelihoods for viral infection are output on the GUI as a graphical scale with a mark placed at a location corresponding to the score. The description also includes a graphical likelihood scale, a mark on the scale, and color-coded GUI elements, together with candidate blood polypeptides including CRP, IP-10, TRAIL, IL1ra, Procalcitonin (PCT), and SAA.

Claims Coverage

Three independent claim features recur across the input items. The claims center on GUI-based computation of a score from blood biomarker values and generation of a graphical likelihood scale for viral infections with a corresponding mark, triggered by a user activation control.

The claimed coverage focuses on user-activated, automated score calculation from blood biomarker values and presenting viral infection likelihoods via a GUI graphical scale with a mark positioned according to the computed score.

Stated Advantages

Documented Applications