Pellionisz, A. (2009) From the Principle of Recursive Genome Function to Interpretation of HoloGenome Regulation by Personal Genome Computers. Cold Spring Harbor Laboratory; Personal Genomes, Sept. 14-17, 2009.
FROM THE PRINCIPLE OF RECURSIVE GENOME FUNCTION TO INTERPRETATION OF HOLOGENOME REGULATION BY PERSONAL GENOME COMPUTERS
Andras J. Pellionisz
HolGenTech, Sunnyvale, California
Recursion is widespread, not excluding genomics; see in quantum theory, in neural network algorithms, recursive PCR and search- & alignment algorithms. By removal of “Junk DNA” and “Central Dogma” axioms, the principle of recursive genome function (Pellionisz, 2008) permits, and the compute-need of analysis of DNA necessitates to supersede dogma and enter a new era of consideration if the intrinsic algorithms of genome function are based on a DNA>RNA>PROTEIN>DNA (etc) recursion.
The intertwined genomic- and epigenomic processes result in the concept of HoloGenome Regulation.
An array of recursive algorithms is surveyed for consideration to interpret recursive genome function: recursion in quantum theory of the HoloGenome, Back-propagation, Tensor Network Theory, Support Vector Machine neural net approaches utilized e.g. in microRNA mining.
Closest attention is focused on “fractal interpretation of genome function” wherein “pyknon-like elements; PLE” (repetitive motifs, Rigoutsos et al., 2006) are the material basis of self-similar repetitions in the DNA, to govern fractal growth (FractoGene, 2002).
New results of an analysis of the smallest full DNA of free living organisms (Mycoplasma Genitalium) where the found PLE frequency over rank log/log curve follows the Zipf-Mandelbrot Parabolic Fractal Distribution Curve (ZMPFDC), while A,C,T,G random-strings of identical length did not show PLE-architecture; ZMPFDC did not appear.
Fractality of DNA provides “algorithmic reference sequence” and thus “fractal defects” can be targeted, reducing the burden on brute force. Deployment of Personal Genome Computers (PGC) accelerates computing, based on FPGA-hybrids eminently suited for small-bit-size string manipulations, developed for defense, financial & graphic computing.
“Fractal defects” correlate with hereditary syndromes. Existing software ported to PGC with hybrid hardware benchmarks acceleration by orders of magnitude.
Verification of the value of human DNA at $5,000 and the value of its interpretation provides a margin of $60,000 making FPGA-hybrid PGC under $20,000 hw and $3,200 sw feasible for the ecosystem.
The singularity of inversion of cost/value takes genome computing to end users, like computers went from monsters of mainframes to widespread and user-friendly PC-s, thereby positively disrupting the ecosystem of computing.
Figures of Dr. Pellionisz' presentation:
Introduction by James Watson
Thomas Hudson, Ontario Institute of Cancer Research
Organizers and Session Chairs:
George Church, Harvard University
Paul Flicek, European Bioinformatics Institute, UK
Richard Gibbs, Baylor College of Medicine
Elaine Mardis, Washington University School of Medicine
C. Thomas Caskey, University of Texas Health Science Center
Presenters at the II. Personal Genomes Conference
in Cold Spring Harbor (September 14-17, 2009):