Evaluation of the Effects of Integer Assignment on the RNA Inverse Folding Problem Using Factorization Machine with Annealing

Evaluation of the Effects of Integer Assignment on the RNA Inverse Folding Problem Using Factorization Machine with Annealing

by Shuta Kikuchi

Formal & Physical Sciences Education JNEducationJNKEducational administration and organization
🎙 Shuta Kikuchi 👥 311 📅 November 28, 2025 ⏱ 25 min 👁 149 📄 original study 🧭 2026-08-16
Available in: English (current) Français

Keywords

RNA inverse foldingFactorization machine with annealingIsing machinesInteger encodingQuantum computing

Summary

Shuta Kikuchi presents a study on applying the Factorization Machine with Annealing (FMA) to the RNA inverse folding problem, which aims to design RNA sequences that fold into a desired secondary structure. The problem is formulated as a combinatorial optimization task, but the objective function is not explicitly defined, so FMA is used as a black-box optimization method. FMA employs a factorization machine as a surrogate model, converting inputs into binary variables. Since RNA bases are categorical without inherent order, they must be encoded as integers, and the choice of encoding may affect performance. The study systematically compares four nucleotide-to-integer assignment schemes and two binary encoding methods (one-hot and domain-wall) on a 27-base target structure from the Eterna benchmark. Results show that FMA successfully generates sequences folding into the target structure, but the choice of assignment significantly influences the minimum free energy and nucleotide composition, especially with domain-wall encoding when the influence of free energy is small. The study extends the applicability of quantum annealing machines to a new domain and highlights the importance of encoding for categorical variables.

179 words

Critical Evaluation

Value of the Information & Strength of the Argument

The value of the information lies in its contribution to applying quantum-inspired optimization to a biological problem, specifically the RNA inverse folding problem. The study is novel in systematically evaluating the effect of integer assignment for categorical variables in FMA, which has not been previously investigated. The argumentation is logical and well-structured: the problem is clearly defined, the methodology is explained, and results are presented with tables and discussion. However, the sample size is limited (27 bases), and the study does not compare with other state-of-the-art methods for RNA inverse folding, which would strengthen the claims. The discussion of why certain assignments lead to biased base frequencies is plausible but not deeply analyzed.

Scientific Rigor, Source Quality, Title Accuracy

The presentation is scientifically rigorous in its methodology, with clear definitions of the objective function and evaluation metrics. However, no specific sources are cited in the talk or description, and the work is not published in a peer-reviewed venue, limiting verifiability. The title accurately reflects the content, and the talk is well-organized. No comments were provided for analysis.

186 words

Title / Content Match

The title accurately reflects the content, focusing on the effect of integer assignment in the RNA inverse folding problem using FMA.

Quality & Reliability

7/10

The presentation is a clear account of original research, with methodology and results presented systematically. However, it is a conference talk with limited detail on validation and no peer-reviewed publication referenced, and the sample size is small (27 bases).

Key Moments

Markers derived by PSI from the transcript: the creator did not define chapters.

Contribution & Novelties

This study contributes to the field by demonstrating the applicability of FMA to the RNA inverse folding problem, a novel application of Ising machines to a biological problem. It systematically evaluates the effect of integer assignment for categorical variables, revealing that the choice of encoding can significantly impact the solutions, particularly with domain-wall encoding. This highlights an unexplored aspect of handling unordered categorical variables in black-box optimization with Ising machines.

Pour aller plus loin :

  • RNA inverse folding problem — Provides background on the problem and its significance.
  • Factorization machines — Overview of the surrogate model used in FMA.
  • Quantum annealing — Context for the Ising machine approach.

108 words

Radar Profile

The radar profile shows high scores in technical level and information quality, indicating a technically detailed and informative presentation. However, the lower scores in reliability and information quantity reflect the limited scope and lack of external validation.

Reliability 6/10