Journal of Modern Classical Physics & Quantum Neuroscience

Abstract

MERP- DNA is a deterministic computational method for transforming DNA sequences into recursive numerical state trajectories. The method encodes A, C, G, and T as 0, 1, 2, and 3 and updates a bounded state using a position-dependent modular recurrence. A sequence of controlled experiments evaluates baseline trajectory generation, single-mutation propagation, mutation burden, mutation position, comparison with Hamming distance, background substitution noise, repeated observations, and formal finite-state-transducer equivalence. The experiments show that a local substitution can propagate through subsequent states and that earlier mutations produce greater trajectory divergence under clean conditions. The clean benchmark also shows that MERP-DNA produces a substantially larger and nonlinear response than Hamming distance, while both methods perfectly classify the simple synthetic mutation/non-mutation benchmark. Under simulated background noise, however, MERP-DNA distances become compressed and can be dominated by noise. Repeated averaging of pairwise MERP distances does not reliably recover the mutation signal under the tested conditions. The paper therefore presents MERP-DNA as a computational bioinformatics proof-of-concept rather than a clinical or diagnostic method. An explicit position-aware deterministic finite-state transducer reproduces the original implementation exactly, providing a formal computer-science interpretation of the machine. Future work should validate the representation using real biological genomic sequences and validated variants.