Central Dogma Transformer III: Interpretable AI Across DNA, RNA, and Protein

• URL: http://arxiv.org/abs/2603.23361v2
• Date: Thu, 26 Mar 2026 16:53:09 GMT
• Title: Central Dogma Transformer III: Interpretable AI Across DNA, RNA, and Protein
• Abstract summary: We present CDT-III, which extends mechanism-oriented AI across the full central dogma: DNA, RNA, and protein.<n>Its two-stage Virtual Cell Embedder architecture mirrors the spatial compartmentalization of the cell.<n>Applying to in silico CD52 knockdown approximating Alemtuzumab, the model predicts 29/29 protein changes correctly.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Biological AI models increasingly predict complex cellular responses, yet their learned representations remain disconnected from the molecular processes they aim to capture. We present CDT-III, which extends mechanism-oriented AI across the full central dogma: DNA, RNA, and protein. Its two-stage Virtual Cell Embedder architecture mirrors the spatial compartmentalization of the cell: VCE-N models transcription in the nucleus and VCE-C models translation in the cytosol. On five held-out genes, CDT-III achieves per-gene RNA r=0.843 and protein r=0.969. Adding protein prediction improves RNA performance (r=0.804 to 0.843), demonstrating that downstream tasks regularize upstream representations. Protein supervision sharpens DNA-level interpretability, increasing CTCF enrichment by 30%. Analysis of experimentally measured mRNA and protein responses reveals that the majority of genes with observable mRNA changes show opposite protein-level changes (66.7% at |log2FC|>0.01, rising to 87.5% at |log2FC|>0.02), exposing a fundamental limitation of RNA-only perturbation models. Despite this pervasive direction discordance, CDT-III correctly predicts both mRNA and protein responses. Applied to in silico CD52 knockdown approximating Alemtuzumab, the model predicts 29/29 protein changes correctly and rediscovers 5 of 7 known clinical side effects without clinical data. Gradient-based side effect profiling requires only unperturbed baseline data (r=0.939), enabling screening of all 2,361 genes without new experiments.

Related papers

• RNA-FM: Flow-Matching Generative Model for Genome-wide RNA-Seq Prediction [48.07329582311906]
  We propose RNA-FM, a flow-matching generative framework for genome-wide bulk RNA-seq prediction from whole-slide images.<n> RNA-FM formulates transcriptomic prediction as a continuous-time conditional transport problem.<n>It enables scalable and biologically interpretable genome-wide gene expression imputation.
  arXiv  Detail & Related papers  (2026-05-12T06:48:58Z)
• Central Dogma Transformer II: An AI Microscope for Understanding Cellular Regulatory Mechanisms [0.0]
  We present CDT-II, an "AI microscope" whose attention maps are directly interpretable as regulatory structure.<n>By mirroring the central dogma in its architecture, CDT-II ensures that each attention mechanism corresponds to a specific biological relationship.<n>Applying to K562 CRISPRi data, CDT-II predicts perturbation effects and recovers the GFI1B regulatory network without supervision.
  arXiv  Detail & Related papers  (2026-02-09T14:54:31Z)
• Central Dogma Transformer: Towards Mechanism-Oriented AI for Cellular Understanding [0.0]
  We present the Central Dogma Transformer (CDT), an architecture that integrates pre-trained language models for DNA, RNA, and protein.<n>We validate CDT v1 on CRISPRi enhancer perturbation data from K562 cells, achieving a Pearson correlation of 0.503.<n>These results suggest that AI architectures aligned with biological information flow can achieve both predictive accuracy and mechanistic interpretability.
  arXiv  Detail & Related papers  (2026-01-03T06:29:22Z)
• Integrated Transcriptomic-proteomic Biomarker Identification for Radiation Response Prediction in Non-small Cell Lung Cancer Cell Lines [7.496897814762568]
  First proteotranscriptomic framework for SF2 prediction in non-small cell lung cancer (NSCLC)<n> integrated transcriptome-proteome framework for identifying concurrent biomarkers predictive of radiation response.<n>Independent pipelines identified 20 prioritized gene signatures from transcriptomic, proteomic, and combined datasets.
  arXiv  Detail & Related papers  (2025-11-27T20:01:51Z)
• TRIDENT: A Trimodal Cascade Generative Framework for Drug and RNA-Conditioned Cellular Morphology Synthesis [56.9460577864211]
  TRIDENT is a cascade generative framework that synthesizes realistic cellular morphology by conditioning on both the perturbation and the corresponding gene expression profile.<n> TRIDENT significantly outperforms state-of-the-art approaches, achieving up to 7-fold improvement with strong generalization to unseen compounds.
  arXiv  Detail & Related papers  (2025-11-23T04:43:27Z)
• DNABERT-2: Fine-Tuning a Genomic Language Model for Colorectal Gene Enhancer Classification [0.0]
  DNABERT-2 is a transformer genomic language model that uses byte-pair encoding to learn variable-length tokens from DNA.<n>Gene enhancers control when and where genes switch on, yet their sequence diversity and tissue specificity make them hard to pinpoint in colorectal cancer.<n>This is the first study to apply a second-generation genomic language model with BPE tokenization to enhancer classification in colorectal cancer.
  arXiv  Detail & Related papers  (2025-09-28T16:10:03Z)
• ProteinZero: Self-Improving Protein Generation via Online Reinforcement Learning [49.2607661375311]
  We present ProteinZero, a novel framework that enables computationally scalable, automated, and continuous self-improvement of the inverse folding model.<n>ProteinZero substantially outperforms existing methods across every key metric in protein design.<n> Notably, the entire RL run on CATH-4.3 can be done with a single 8 X GPU node in under 3 days, including reward.
  arXiv  Detail & Related papers  (2025-06-09T06:08:59Z)
• scHyena: Foundation Model for Full-Length Single-Cell RNA-Seq Analysis in Brain [46.39828178736219]
  We introduce scHyena, a foundation model designed to address these challenges and enhance the accuracy of scRNA-seq analysis in the brain. scHyena is equipped with a linear adaptor layer, the positional encoding via gene-embedding, and a bidirectional Hyena operator. This enables us to process full-length scRNA-seq data without losing any information from the raw data.
  arXiv  Detail & Related papers  (2023-10-04T10:30:08Z)
• hist2RNA: An efficient deep learning architecture to predict gene expression from breast cancer histopathology images [11.822321981275232]
  Deep learning algorithms can effectively extract morphological patterns in digital histopathology images to predict molecular phenotypes quickly and cost-effectively. We propose a new, computationally efficient approach called hist2RNA inspired by bulk RNA-sequencing techniques to predict the expression of 138 genes.
  arXiv  Detail & Related papers  (2023-04-10T10:54:32Z)
• Knowledge from Large-Scale Protein Contact Prediction Models Can Be Transferred to the Data-Scarce RNA Contact Prediction Task [40.051834115537474]
  We find that a protein-coevolution Transformer-based deep neural network can be transferred to the RNA contact prediction task. Experiments confirm that RNA contact prediction through transfer learning is greatly improved. Our findings indicate that the learned structural patterns of proteins can be transferred to RNAs, opening up potential new avenues for research.
  arXiv  Detail & Related papers  (2023-02-13T06:00:56Z)
• Structure-informed Language Models Are Protein Designers [69.70134899296912]
  We present LM-Design, a generic approach to reprogramming sequence-based protein language models (pLMs) We conduct a structural surgery on pLMs, where a lightweight structural adapter is implanted into pLMs and endows it with structural awareness. Experiments show that our approach outperforms the state-of-the-art methods by a large margin.
  arXiv  Detail & Related papers  (2023-02-03T10:49:52Z)
• State-specific protein-ligand complex structure prediction with a multi-scale deep generative model [68.28309982199902]
  We present NeuralPLexer, a computational approach that can directly predict protein-ligand complex structures. Our study suggests that a data-driven approach can capture the structural cooperativity between proteins and small molecules, showing promise in accelerating the design of enzymes, drug molecules, and beyond.
  arXiv  Detail & Related papers  (2022-09-30T01:46:38Z)
• Accurate RNA 3D structure prediction using a language model-based deep learning approach [50.193512039121984]
  RhoFold+ is an RNA language model-based deep learning method that accurately predicts 3D structures of single-chain RNAs from sequences.<n>RhoFold+ offers a fully automated end-to-end pipeline for RNA 3D structure prediction.
  arXiv  Detail & Related papers  (2022-07-04T17:15:35Z)
• Predicting Hydroxyl Mediated Nucleophilic Degradation and Molecular Stability of RNA Sequences through the Application of Deep Learning Methods [0.0]
  This paper proposes and evaluates three deep learning models as methods to predict the reactivity and risk of degradation of mRNA sequences. The Stanford Open Vaccine dataset of 6034 mRNA sequences was used in this study. Results suggest these models can be applied to understand and predict the chemical stability of mRNA in the near future.
  arXiv  Detail & Related papers  (2020-11-09T10:42:53Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.