White Paper

Helixometry, Inc.

Personalized Medicine Intellectual Portfolio
with FractoGene Patent Pending for mathematics of DNA


Dr. András J. Pellionisz

Founder and CEO of HelixoMetry, Inc.

formerly Chief Intelligence Officer, Senior Knowledge Engineer, Sr. Software Architect of
Local and Global Silicon Valley Information Technology Companies,
Senior National Research Council Associate of the National Academy to NASA,
Professor of former Department of Physiology and Biophysics, New York University

(408) 891.7187
cob@helixometry.com


Introduction

The "Double Helix" is not even half a Century "old". Now after the Human Genome Project that revealed an entire strand of the genetic material of the human, the December 5th, 2002 issue of Nature also made public the entire DNA information of a second species (the mouse).

This, by all means, represents a new Chapter in the Genome Revolution, directly aimed at "cracking the code" with the help of a) comparing the human and mouse genome, b) enabling directly comparable experimentation that of course can not be done on humans.

Therefore, HelixoMetry's big leap in the Intellectual Portfolio, adding FractoGene Patent (Pending) by mid-summer of 2003 was well timed. It enables HelixoMetry to enter the head-on competition with the advantage of proprietary utility to measure and compare the introns by fractal geometrical (mathematical) methods, while others are still debating if introns are "junk DNA" or maybe not.

Until December 5th, the main challenge for Genomics was to use Pattern Recognition, a well-established field for several decades, for samples from just one entire set of the Genome.

AGILENT summarizes, in one diagram, the challenge dominating the first chapter of Personalized Medicine - that is microarray pattern matching, related to the ultimate personalization, the double helix.

Genomics is one of the most crucial emerging fields of application for "Neural Nets" (see GRAIL), with the leading Neurocomputing firms positioning to Genomics. Of course, "rules" are already used (AI). "Tags" are also of use, for instance to mark on which chromosome does a gene take place, see "XML" for taxonomies of genes and even species. Natural Language Processing is also used, to mine vast medical history data banks.

Dr. Pellionisz, Founder and CEO of Helixometry, elaborates his accomplishments elsewhere. Suffice to point out here just a few of the landmarks of his "third of a Century pioneering in Pattern Recognition in Biological Systems with Neural Nets, AI, NLP and XML". As for the FractoGene Patent (Pending), the entire domain http://fractogene.com is devoted to the subject.

I. From Pattern Analysis of Neural Nets to Information Geometry of the Biological Brain (1967-1989)

1969 is the date of a seminal article (first computer analysis of Patterns in specific Biological Neural Nets). Analysis starts from biological reality of the "little brain" (cerebellum), and uses "large scale" simulations of pattern transfer pattern by computers (34 years ago, in Hungary).

1973, less than half a decade, working closely with experimental neuroscientist Prof. János Szentágothai, Pellionisz' computer analysis of such complex biological patterns, reminiscent to microarray patterns, becomes an established methodology accepted by most reputable referenced journals.

1985 Pellionisz, as a Professor at former Department of Physiology and Biophysics of New York University, working closely with experimental neuroscientist Prof. Rodolfo Llinás, works out the neurocomputer algorithms and computer analysis into an information geometry approach. Prof. Patricia Churchland (San Diego) compares this Tensor Network Theory with Brain Theory of Francis Crick (MIT Press: Neurophilosophy, first edition 1986, paperback 1989)

1985 was a landmark in quantitatively elaborating the mapping of one multidimensional biological reality (vestibular apparatus) onto another multidimensional biological reality (oculomotor apparatus). These multiple mappings by neural nets, help solve the problem how the double helix information is mapped on genome, protein, gene expression profile, microarray pattern and finally drug space, as shown in the "concept diagram" of Helixometry.

1986 in the American Institute of Physics (see neural network pattern diagram), the key to neural net patterns is that they express a geometrical relationship of biological realities of the complex gaze system.

1987, less than two decades of work, led Tensor Network Theory pioneering into the "Encyclopedia of Neuroscience", having worked out tensor theory, the intrinsic mathematics of neural patterns, and thus explaining 1/3 of the brain (the cerebellum). The mathematical theory yielded quantitative predictions, verified by independent experimentalists.

1990 Information Geometry approach of Dr. Pellionisz is published in the "MIT Neural Net bestseller" - and a lucid, "easy to digest" 5-page summary is written by Prof. Anderson. (1,2,3,4,5). Pellionisz' TNT receives Humboldt Prize from Germany, the first international research prize in Neurocomputing; Hopfield and Hecht-Nielsen receive USA awards.


II. Large-Scale Information Technology Applications of Neural Nets at NASA (1990-1999)

1990 By invitation of NASA, a Neural Net Program is elaborated for large-scale applications, such as Flight Control by Neural Nets (see NASA-NIHM-NSF program, concept-diagram, concept-application, NASA-application, NASA-design, NASA-plan till 96, NASA-team, a NASA $8 M budget for this pilot-project). Program also suggests Genome Project for NASA Ames Information Technology Branch, Biocomputing Center. Paul Werbos, NSF Neurocomputing chief and inventor of biologically impossible "backpropagation" algorithm, supports convergence of industrial applications based on biological realism.

1991 Industrialization of Neural Nets is made difficult if Geometries are grossly non-Euclidean, or the underlying geometry is fractal. It is shown that there is not enough genetic information to determine brain cells without a fractal information-compression.

1999 Initiated in 1990 by Pellionisz, NASA implements industrial-scale Project, flying an F15 by Neurocomputer Geometrical Reconfiguration.


III. Large-Scale Information Technology Applications for Internet Personalization (1994-2001)

1994 (111-112) "From Geometrical Foundations of NN Research to Lead-Roles in Silicon Valley Information Industry" - Silicon Valley Net Institute is formed to integrate heterogeneous Information Technologies of Rule Based (AI), Tag (XML), Natural Language Processing (NLP) and Neural Net (NN) layers into SVNI's proprietary 4-layer IT Architecture.

1997-2001 Executive of Ernst & Young software house (Kanisa) and Chief Intelligence Officer of Silicon Valley local and Global Internet companies (Verge, Xmarksthespot, Mindmaker) to accumulate industrial experience in applying Personalization Information Technology.

2000-2001 expressed in the ZDNET video-interview given by Dr. Pellionisz to Peter Coffee at COMDEX-2000, personalization is a major trend in the 21st Century.

IV. HELIXOMETRY: Large-Scale Information Technology Applications of Neural Nets and FractoGene Patent Pending, for Personalized Medicine (2002-2003)

2002 Helixometry is incorporated in California and applies the Information Technology and Intellectual Property of Personalization acquired, fully owned and tested and proven in e-commerce to Personalized Medicine.

Business opportunity is based on this proven technology-base and newly introduced Intellectual Property Portfolio with the key FractoGene Patent (Pending) that provides a mathematical (fractal geometrical) language for the genetic code.

Helixometry agrees with IBM, that sees Personalized Medicine as a $43 Billion market by 2005 for IT, in the conclusion that in addition to "Big Pharma", "Big IT" will play a dominant role in Personalized Medicine.

Marketing strategy and Helixometry Financing of the HelixoMetry Intellectual Property Portfolio is elaborated elsewhere.