Reverse Central Dogma: Computational Exploration of Protein-to-Nucleic Acid Information Flow

This seminar presentation synthesizes two interconnected research initiatives that together challenge conventional paradigms in molecular biology and drug discovery through advanced computational methodologies. The first project, "Reverse Central Dogma," confronts the foundational principle of molecular biology by demonstrating the theoretical feasibility of reverse information flow from proteins to nucleic acids. Through extensive molecular docking simulations involving 12 distinct peptide motifs and 4 nucleotide types, we identified statistically significant amino acid-nucleotide binding specificities, with arginine-rich peptides exhibiting exceptional preference for guanine nucleotides (ΔG = -9.8 kcal/mol, p < 0.01). We developed and validated a novel reverse translation algorithm that successfully converts protein sequences into probable nucleic acid sequences with confidence levels of 72–83%, employing Boltzmann probability distributions and Monte Carlo sampling. This work establishes a robust theoretical foundation for peptide-guided nucleic acid assembly, with profound implications for origin-of-life scenarios and synthetic biology applications. The second project, "Seed to Lead," translates computational innovation into therapeutic potential by developing a mechanism-driven pipeline to valorize Nepalese hemp seed byproducts for cancer therapy. This research focuses on underutilized lignanamides—bioactive compounds from Cannabis seeds—and employs a three-phase computational-experimental strategy. Through in silico molecular docking, we identified promising lignanamide interactions with Nicotinamide Phosphoribosyltransferase (NAMPT), a critical enzyme in cancer metabolism. Our designed lead compound, CSE-03, demonstrates predicted moderate inhibition with high selectivity (Therapeutic Index: 3.8), addressing the historical failure of potent NAMPT inhibitors due to systemic toxicity. The project integrates phytochemical extraction, ADMET prediction, and in vitro validation to create a cost-effective blueprint for mechanism-guided drug discovery in resource-limited settings. Collectively, this presentation illustrates how computational biology serves as a unifying bridge between theoretical innovation and applied therapeutic development. By employing shared methodologies; molecular docking, statistical modeling, and algorithm development, we simultaneously challenge six decades of biological dogma and propose novel solutions to twenty-year clinical failures. This work not only advances our fundamental understanding of biological information systems but also demonstrates a practical pathway for leveraging local biodiversity to address global health challenges, exemplifying the transformative potential of integrated computational-experimental approaches in modern biotechnology.