Regressor-guided Diffusion Model for De Novo Peptide Sequencing with Explicit Mass Control

Shaorong Chen; Jingbo Zhou; Jun Xia

q-bio.QMFeb 23, 2026

Regressor-guided Diffusion Model for De Novo Peptide Sequencing with Explicit Mass Control

Authors: Shaorong Chen, Jingbo Zhou, Jun Xia

Organizations: Zhejiang University, Hangzhou, China, 310058 · AI Lab, Westlake University, Hangzhou, China, 310030 · AIMS Lab, The Hong Kong University of Science and Technology (Guangzhou), Guangzhou, China, 511453 · The Hong Kong University of Science and Technology, Hong Kong, China, 999077

Abstract

The discovery of novel proteins relies on sensitive protein identification, for which de novo peptide sequencing (DNPS) from mass spectra is a crucial approach. While deep learning has advanced DNPS, existing models inadequately enforce the fundamental mass consistency constraint, that a predicted peptide's mass must match the experimental measured precursor mass. Previous DNPS methods often treat this critical information as a simple input feature or use it in post-processing, leading to numerous implausible predictions that do not adhere to this fundamental physical property. To address this limitation, we introduce DiffuNovo, a novel regressor-guided diffusion model for de novo peptide sequencing that provides explicit peptide-level mass control. Our approach integrates the mass constraint at two critical stages: during training, a novel peptide-level mass loss guides model optimization, while at inference, regressor-based guidance from gradient-based updates in the latent space steers the generation to compel the predicted peptide adheres to the mass constraint. Comprehensive evaluations on established benchmarks demonstrate that DiffuNovo surpasses state-of-the-art methods in DNPS accuracy. Additionally, as the first DNPS model to employ a diffusion model as its core backbone, DiffuNovo leverages the powerful controllability of diffusion architecture and achieves a significant reduction in mass error, thereby producing much more physically plausible peptides. These innovations represent a substantial advancement toward robust and broadly applicable DNPS. The source code is available in the supplementary material.

Figures & tables

Figure 1: Schematic of a typical bottom-up proteomics workflow ( Zhang et al. 2013 ) . Proteins are digested into peptides and analyzed by mass spectrometer to produce mass spectrum. DNPS is the process of inferring a peptide’s sequence directly from its spectrum. The resulting peptide sequences can then be used for protein sequence assembly.

Panel of Figure 2 — Figure 2: Performance comparison of DiffuNovo with state-of-the-art methods on mass-related metrics. Plausible prediction rate is the percentage of predicted peptides whose mass satisfies the experimental mass (higher is better). Mass error is the mean absolute error between the predicted peptide mass and the experimental mass (lower is better).

Figure 3: The architecture of DiffuNovo. In the Forward Diffusion Process, a ground-truth peptide sequence is converted into an embedding $z_{0}$ and then corrupted by adding noise over $T$ timesteps to produce a noisy latent variable $z_{T}$ . DiffuNovo is trained to reverse this process. The Spectrum Encoder transforms the mass spectrum into embedding $\mathbf{x}$ . The core of DiffuNovo is Reverse Diffusion Process where the Peptide Decoder denoise the noisy latent $z_{t}$ into a cleaner latent $z_{t-1}$ conditioned on $\mathbf{x}$ . Critically, the Peptide Mass Regressor provides guidance to the Peptide Decoder at each timestep $t$ , compelling the prediction adheres to the mass constraint. Finally, the clean latent $\hat{z}_{0}$ is passed to a prediction head to output the final predicted peptide.

z_{0}\sim q_{\phi}(z_{0}\mid\mathbf{y})=\mathcal{N}(z_{0};\mathrm{Emb}_{\phi}(\mathbf{y}),(1-\alpha_{0})\mathbf{I})

(12)

Algorithm 1 Training Stage of DiffuNovo.

{\mu}_{\theta}(z_{t_{i}},t_{i},\mathbf{x})\leftarrow\lambda_{1}z_{t_{i}}+\lambda_{2}\mathbf{g}_{\theta}(z_{t_{i}},t_{i},\mathbf{x})

(15)

Algorithm 2 Inference Process of DiffuNovo.

Panel of Figure 4 — Figure 4: Performance comparison on mass-related metrics.

Method	HelixNovo	CasaNovo	DiffuNovo
MAE( $\downarrow$ )	21.388	8.583	2.251
Analysis of DiffuNovo’s Improvement
DiffuNovo vs.	HelixNovo	CasaNovo	DiffuNovo
X% Decrease	89.5%	73.8%	0%
Fold Decrease	9.50 $\times$	3.81 $\times$	1.00 $\times$

Table 1: Evaluation of mass error. We compares DiffuNovo with two baselines and reports the Mean Absolute Error (MAE), where lower is better. The upper section shows the mass errors. The lower section quantifies this improvement.

	Peptide-level Performance						Amino Acid-level Performance
Models	9-species		HC-PT		7-species		9-species		HC-PT		7 species
	Prec.	AUC	Prec.	AUC	Prec.	AUC	Prec.	Recall	Prec.	Recall	Prec.	Recall
DeepNovo	0.428	0.376	0.313	0.255	0.204	0.136	0.696	0.638	0.531	0.534	0.492	0.433
PointNovo	0.480	0.436	0.419	0.373	0.022	0.007	0.740	0.671	0.623	0.622	0.196	0.169
CasaNovo	0.481	0.439	0.211	0.177	0.119	0.084	0.697	0.696	0.442	0.453	0.322	0.327
AdaNovo	0.505	0.469	0.212	0.178	0.174	0.135	0.698	0.709	0.442	0.451	0.379	0.385

Table 2: The comparison of de novo peptide sequencing performance between our proposed model, DiffuNovo, and other state-of-the-art methods on the three benchmark datasets. We report precision and AUC at the peptide level, and precision and recall at the amino acid level. The best and the second best are highlighted with bold and underline , respectively.

	PTM Precision
Models	Nine-Species	HC-PT	Seven-Species
DeepNovo	0.576	0.626	0.391
PointNovo	0.629	0.676	0.117
AdaNovo	0.652	0.552	0.448
CasaNovo	0.706	0.501	0.360
HelixNovo	0.680	0.568	0.473

Table 3: Empirical comparison of PTM identification. We evaluate the ability of DiffuNovo and other models to identify Post-Translational Modifications (PTMs) on the benchmark datasets. The best results and the second best are highlighted with bold and underline , respectively.

Figure 5: Mass error loss curves. The figure illustrates the mass error loss (MSE) as a function of the normalized diffusion timestep. We conducted experiment on 5 random batches (2560 samples), calculating and plotting the loss for each batch (dashed line) and the average loss (solid line).

Model	Performance		Mass Constraint
	Peptide	AA	Plausible	MAE
	Prec.( $\uparrow$ )	Prec.( $\uparrow$ )	Rate( $\uparrow$ )	Value( $\downarrow$ )
Full Model	0.485	0.648	0.843	4.350
- w/o Encoder	0.104	0.313	0.686	16.509
- w/o Decoder	0.066	0.267	0.652	35.260
- w/o Regressor	0.437	0.641	0.703	14.555

Table 4: Ablation study of the model components.

Mass Spectrometry

Mass spectrometry measures ion mass-to-charge patterns to identify and quantify molecules, with AI linking spectra to structures, peptides, and biological signals.

See all 33 papers

MemNovo: Look Back at the Spectrum for Balanced De Novo Peptide Sequencing from Mass Spectrometry
Jun 10, 2026Dongxin Lyu, Jingbo Zhou, Hongxin Xiang +2Memory-Augmented Neural Networks Transformer Inference
Westlake University Hangzhou, Zhejiang, China · Hunan University Changsha, Hunan, China · Shanghai Artificial Intelligence Laboratory Shanghai, China +1
GyroNovo: Error-Guided Fragment Imputation with Mass-Aware Attention for De Novo Peptide Sequencing
Sep 24, 2026Abdellah El Mekki, Laks V. S. Lakshmanan, Muhammad Abdul-MageedMass Spectrometry Attention Mechanisms
The University of British Columbia · Canada Research Chair in NLP and ML
dIon: Fragmentation-Based Invariance for Self-Supervised Learning of Tandem Mass Spectra
Oct 5, 2026Alfred Nilsson, Joel Lapin, Samuel H. Payne +2Self-Supervised Pre-Training Mass Spectrometry
Science for Life Laboratory, KTH Royal Institute of Technology Stockholm, Sweden · Computational Mass Spectrometry, TUM School of Life Sciences Technical University of Munich, Freising, Germany · Biology Department, Brigham Young University Provo, Utah 84602, United States +1

Jun 10, 2026J/K move · Enter open · S save

MemNovo: Look Back at the Spectrum for Balanced De Novo Peptide Sequencing from Mass Spectrometry

Dongxin Lyu, Jingbo Zhou, Hongxin Xiang, Yuqiang Li+1

Westlake University Hangzhou, Zhejiang, China · Hunan University Changsha, Hunan, China · Shanghai Artificial Intelligence Laboratory Shanghai, China+1

Memory-Augmented Neural Networks Transformer Inference Mass Spectrometry Autoregressive Decoding

Open paper AskPDF arXiv

De novo peptide sequencing from tandem mass spectrometry is pivotal in proteomics, enabling identification of novel peptides without reference databases. While recent Transformer-based encoder-decoder models have achieved remarkable performance, we uncover a critical pathology in their inference dynamics. Through comprehensive feature scaling experiments, we demonstrate that existing auto-regressive peptide decoders tend to over-rely on generated-sequence priors while progressively under-utilizing fine-grained physical evidence from the input mass spectrum. This phenomenon leads to suboptimal results, where generated peptide sequences are biologically plausible yet not faithful to the input spectrum. To rectify this, we propose MemNovo, a training-free and plug-and-play mechanism that re-balances peptide and spectral contributions at inference time. MemNovo alleviates the information bottleneck by establishing a persistent spectral memory bank and injecting retrieved features directly into the final decoding stage via an ultra-conservative residual connection. Theoretical analysis confirms that this mechanism restores the mutual information between the decoder state and the raw spectrum. Extensive experiments on the Nine Species benchmark with two representative baselines, Casanovo and InstaNovo, demonstrate that MemNovo consistently improves both amino acid precision and peptide precision, achieving up to 39.1% relative improvement in peptide precision for Casanovo and up to 3.9% for InstaNovo, with negligible computational overhead.

Abstract

Mass Spectrometry

Explore similar work

MemNovo: Look Back at the Spectrum for Balanced De Novo Peptide Sequencing from Mass Spectrometry

GyroNovo: Error-Guided Fragment Imputation with Mass-Aware Attention for De Novo Peptide Sequencing

dIon: Fragmentation-Based Invariance for Self-Supervised Learning of Tandem Mass Spectra