Research progress and application of retention time prediction method based on deep learning

Zhuokun Du, Wei Shao, Weijie QinChinese 2021 · DOI

Abstract

In "shotgun" proteomics, the identification of peptides from tryptic digests relies heavily on liquid chromatography-mass spectrometry (LC-MS). A critical component of this process is the retention time (RT)—the duration a peptide remains within a chromatographic column. While traditional prediction methods relied on manual calculations of physicochemical properties, they often lacked the flexibility to adapt across different laboratory conditions or instrument platforms. Recently, deep learning (DL) has emerged as a transformative solution, leveraging multi-layered neural networks to learn complex peptide-RT relationships directly from raw sequence data. This paper reviews the progress of DL-based RT prediction, emphasizing the role of transfer learning in cross-condition calibration and the development of pseudo-spectral libraries for data-independent acquisition (DIA). Despite challenges in predicting complex modifications like glycosylation, DL methods have significantly improved the accuracy of peptide identification and quality control in proteomics research.

Introduction

The "shotgun" proteomics strategy is the cornerstone of modern protein analysis, where proteins are enzymatically digested into peptides and separated via liquid chromatography before mass spectrometry analysis. In this workflow, the retention time (RT) serves as a vital "fingerprint" for peptide identification. Historically, researchers utilized Quantitative Structure-Retention Relationship (QSRR) models to predict RT. However, these models required extensive manual feature engineering—such as calculating hydrophobicity or helicity—and were notoriously rigid, often failing when applied to different chromatographic conditions or laboratory setups.

The advent of deep learning has shifted this paradigm. By automating feature extraction and utilizing large-scale datasets, deep learning models provide superior accuracy and adaptability. This review explores the architectures, encoding strategies, and practical applications of these models in contemporary proteomics.

Methodological Framework

Deep learning models have evolved from simple architectures to complex networks capable of capturing the nuances of peptide chemistry:

  • Architectures: Modern methods utilize Convolutional Neural Networks (CNNs) to detect local amino acid interactions and Recurrent Neural Networks (RNNs), specifically Bidirectional Long Short-Term Memory (BiLSTM) units, to capture long-range dependencies within a sequence. More advanced models have even integrated Capsule Networks (CapsNet) to model hierarchical relationships.
  • Sequence Encoding: To process peptide information, models use various encoding techniques. These range from simple one-hot encoding to dense embeddings that represent chemical properties. Some models employ atom-level encoding (counting C, H, N, O, P, and S atoms) to better account for chemical modifications.
  • Transfer Learning: This is perhaps the most significant advantage of deep learning in this field. Models like Prosit and DeepRT are pre-trained on massive datasets (e.g., ProteomeTools). Researchers can then "fine-tune" these models on small, lab-specific datasets, allowing the model to adapt to new chromatography conditions without requiring a complete retrain.

Applications in Modern Proteomics

The integration of high-accuracy RT prediction has led to two major breakthroughs in the field:

1. Pseudo-Spectral Libraries for DIA

Data-Independent Acquisition (DIA) typically requires a pre-existing spectral library generated from Data-Dependent Acquisition (DDA) experiments. This is time-consuming and limits discovery to previously observed peptides. By combining predicted RT with predicted fragment ion intensities, researchers can generate "pseudo-spectral libraries." These theoretical libraries allow for "library-free" DIA analysis, enabling the identification of peptides that were never captured in a pilot DDA experiment.

2. Quality Control and PSM Validation

The difference between the predicted RT and the observed RT (known as "Delta RT") serves as an unbiased metric for evaluating the quality of peptide-spectrum matches (PSMs). High-accuracy predictions allow researchers to filter out false positives, ensuring that the reported identifications are statistically sound.

Key Results and Performance

Deep learning models have consistently outperformed traditional tools like SSRCalc. For instance, the Prosit model achieved a near-perfect correlation ($R = 1.00$) on synthetic peptide sets, with a 95% confidence interval of only 4.25 iRT units. Furthermore, the use of predicted libraries in workflows like DeepMass:Drip or DeepDIA has been shown to increase the number of identified peptides by 20% to 31% compared to standard experimental libraries. Transfer learning has also enabled models to maintain high accuracy ($R^2 > 0.97$) across diverse chromatography modes, including RPLC, SCX, and HILIC.

Challenges and Future Outlook

While deep learning has revolutionized RT prediction, several hurdles remain. Predicting RT for complex modifications, such as glycosylation, is difficult due to the structural diversity of sugar chains and a lack of large-scale training data. Additionally, most current models treat peptides as linear strings; incorporating 3D structural context or more sophisticated atomic modeling remains an active area of research.

The ultimate goal is the development of a truly universal model that requires zero calibration across different instruments. As datasets continue to grow and architectures become more refined, deep learning will remain at the forefront of proteomics, turning retention time from a variable nuisance into a powerful tool for discovery.

Summary updated

Abstract

<p id=C2>In “shotgun” proteomics strategy, the proteome is explained by analyzing tryptic digested peptides using liquid chromatography-mass spectrometry. In this strategy, the retention time of peptides in liquid chromatography separation can be predicted based on the peptide sequence. This is a useful feature for peptide identification. Therefore, the prediction of the retention time has attracted much research attention. Traditional methods calculate the physical and chemical properties of the peptides based on their amino acid sequence to obtain the retention time under certain chromatography conditions; however, these methods cannot be directly adopted for other chromatography conditions, nor can they be used across laboratories or instrument platforms. To solve this problem, in recent years, deep learning was introduced to proteomics research for retention time prediction. Deep learning is an advanced machine-learning method that has extraordinary capability to learn complex relationships from large-scale data. By stacking multiple hidden neural networks, deep learning can ingest raw data without manually designed features. Transfer learning is an important method in deep learning. It improves the learning process a new task through the transfer of knowledge from an already-learned related task. Transfer learning allows models trained using large datasets to be utilized across conditions by fine-tuning on smaller datasets, instead of retraining the whole model. Many retention time prediction methods have been developed. In the process of training the model, the sequences of peptides are encoded to represent peptide information. Deep learning considers the relationship between the characteristics of the peptides and their corresponding retention times without the need for manual input of the physical and chemical properties of the peptides. Compared with traditional methods, deep learning methods have higher accuracy and can be easily used under different chromatography conditions by transfer learning. If there are not enough datasets to train a new model, a trained model from other datasets can be used as a replacement after calibration with small datasets obtained from these chromatography conditions. While the retention times of modified peptides can also be predicted, the predictions are inadequate for complex modifications such as glycosylation, and this is one of the main problems to be solved. The predicted retention times were used to control the quality of peptide identification. With high accuracy, the predicted retention times can be considered as actual retention times. Therefore, the difference between predicted and observed retention times can serve as an effective and unbiased quantitative metric for evaluating the quality of peptide-spectrum matches (PSMs) reported using different peptide identification methods. Combined with fragment ion intensity prediction, retention time prediction is used to generate spectral libraries for data-independent acquisition (DIA)-based mass spectrometry analysis. Generally, DIA methods identify peptides using specific spectrum libraries obtained from data-dependent acquisition (DDA) experiments. As a result, only peptides detected in the DDA experiments can be present in the libraries and detected in DIA. Furthermore, it takes a lot of time and effort to build libraries from DDA experiments, and typically, they cannot be adopted across different laboratories or instrument platforms. In contrast, the pseudo spectral libraries generated by retention times and fragment ion intensity prediction can overcome these shortcomings. The pseudo spectral libraries generate theoretical spectra of all possible peptides without the need for DDA experiments. This paper reviews the research progress of deep learning methods in the prediction of retention time and in related applications in order to provide references for retention time prediction and protein identification. At the same time, the development direction and application trend of retention time prediction methods based on deep learning are discussed.
BibTeX
@inproceedings{du2021research,
  author = {Zhuokun Du and Wei Shao and Weijie Qin},
  title = {{Research progress and application of retention time prediction method based on deep learning}},
  booktitle = {Chinese},
  year = {2021},
  doi = {10.3724/sp.j.1123.2020.08015},
}
Figure