Test-Time Training for Deep MS/MS Spectrum Prediction Improves Peptide Identification.

In bottom-up proteomics, peptide-spectrum matching is critical for peptide and protein identification. Recently, deep learning models have been used to predict tandem mass spectra of peptides, enabling the calculation of similarity scores between the predicted and experimental spectra for peptide-spectrum matching. These models follow the supervised learning paradigm, which trains a general model using paired peptides and spectra from standard data sets and directly employs the model on experimental data. However, this approach can lead to inaccurate predictions due to differences between the training data and the experimental data, such as sample types, enzyme specificity, and instrument calibration. To tackle this problem, we developed a test-time training paradigm that adapts the pretrained model to generate experimental data-specific models, namely, PepT3. PepT3 yields a 10-40% increase in peptide identification depending on the variability in training and experimental data. Intriguingly, when applied to a patient-derived immunopeptidomic sample, PepT3 increases the identification of tumor-specific immunopeptide candidates by 60%. Two-thirds of the newly identified candidates are predicted to bind to the patient's human leukocyte antigen isoforms. To facilitate access of the model and all the results, we have archived all the intermediate files in Zenodo.org with identifier 8231084.

Learning to Double-Check Model Prediction From a Causal Perspective.

The present machine learning schema typically uses a one-pass model inference (e.g., forward propagation) to make predictions in the testing phase. It is inherently different from human students who double-check the answer during examinations especially when the confidence is low. To bridge this gap, we propose a learning to double-check (L2D) framework, which formulates double check as a learnable procedure with two core operations: recognizing unreliable predictions and revising predictions. To judge the correctness of a prediction, we resort to counterfactual faithfulness in causal theory and design a contrastive faithfulness measure. In particular, L2D generates counterfactual features by imagining: "what would the sample features be if its label was the predicted class" and judges the prediction by the faithfulness of the counterfactual features. Furthermore, we design a simple and effective revision module to revise the original model prediction according to the faithfulness. We apply the L2D framework to three classification models and conduct experiments on two public datasets for image classification, validating the effectiveness of L2D in prediction correctness judgment and revision.

Reinforced Causal Explainer for Graph Neural Networks.

Explainability is crucial for probing graph neural networks (GNNs), answering questions like "Why the GNN model makes a certain prediction?". Feature attribution is a prevalent technique of highlighting the explanatory subgraph in the input graph, which plausibly leads the GNN model to make its prediction. Various attribution methods have been proposed to exploit gradient-like or attention scores as the attributions of edges, then select the salient edges with top attribution scores as the explanation. However, most of these works make an untenable assumption - the selected edges are linearly independent - thus leaving the dependencies among edges largely unexplored, especially their coalition effect. We demonstrate unambiguous drawbacks of this assumption - making the explanatory subgraph unfaithful and verbose. To address this challenge, we propose a reinforcement learning agent, Reinforced Causal Explainer (RC-Explainer). It frames the explanation task as a sequential decision process - an explanatory subgraph is successively constructed by adding a salient edge to connect the previously selected subgraph. Technically, its policy network predicts the action of edge addition, and gets a reward that quantifies the action's causal effect on the prediction. Such reward accounts for the dependency of the newly-added edge and the previously-added edges, thus reflecting whether they collaborate together and form a coalition to pursue better explanations. It is trained via policy gradient to optimize the reward stream of edge sequences. As such, RC-Explainer is able to generate faithful and concise explanations, and has a better generalization power to unseen graphs. When explaining different GNNs on three graph classification datasets, RC-Explainer achieves better or comparable performance to state-of-the-art approaches w.r.t. two quantitative metrics: predictive accuracy, contrastivity, and safely passes sanity checks and visual inspections. Codes and datasets are available at https://github.com/xiangwang1223/reinforced_causal_explainer.

Personalized Latent Structure Learning for Recommendation.

In recommender systems, users' behavior data are driven by the interactions of user-item latent factors. To improve recommendation effectiveness and robustness, recent advances focus on latent factor disentanglement via variational inference. Despite significant progress, uncovering the underlying interactions, i.e., dependencies of latent factors, remains largely neglected by the literature. To bridge the gap, we investigate the joint disentanglement of user-item latent factors and the dependencies between them, namely latent structure learning. We propose to analyze the problem from the causal perspective, where a latent structure should ideally reproduce observational interaction data, and satisfy the structure acyclicity and dependency constraints, i.e., causal prerequisites. We further identify the recommendation-specific challenges for latent structure learning, i.e., the subjective nature of users' minds and the inaccessibility of private/sensitive user factors causing universally learned latent structure to be suboptimal for individuals. To address these challenges, we propose the personalized latent structure learning framework for recommendation, namely PlanRec, which incorporates 1) differentiable Reconstruction, Dependency, and Acyclicity regularizations to satisfy the causal prerequisites; 2) Personalized Structure Learning (PSL) which personalizes the universally learned dependencies through probabilistic modeling; and 3) uncertainty estimation which explicitly measures the uncertainty of structure personalization, and adaptively balances personalization and shared knowledge for different users. We conduct extensive experiments on two public benchmark datasets from MovieLens and Amazon, and a large-scale industrial dataset from Alipay. Empirical studies validate that PlanRec discovers effective shared/personalized structures, and successfully balances shared knowledge and personalization via rational uncertainty estimation.

Causal Incremental Graph Convolution for Recommender System Retraining.

The real-world recommender system needs to be regularly retrained to keep with the new data. In this work, we consider how to efficiently retrain graph convolution network (GCN)-based recommender models that are state-of-the-art techniques for the collaborative recommendation. To pursue high efficiency, we set the target as using only new data for model updating, meanwhile not sacrificing the recommendation accuracy compared with full model retraining. This is nontrivial to achieve since the interaction data participates in both the graph structure for model construction and the loss function for model learning, whereas the old graph structure is not allowed to use in model updating. Toward the goal, we propose a causal incremental graph convolution (IGC) approach, which consists of two new operators named IGC and colliding effect distillation (CED) to estimate the output of full graph convolution. In particular, we devise simple and effective modules for IGC to ingeniously combine the old representations and the incremental graph and effectively fuse the long- and short-term preference signals. CED aims to avoid the out-of-date issue of inactive nodes that are not in the incremental graph, which connects the new data with inactive nodes through causal inference. In particular, CED estimates the causal effect of new data on the representation of inactive nodes through the control of their collider. Extensive experiments on three real-world datasets demonstrate both accuracy gains and significant speed-ups over the existing retraining mechanism.

Folic acid-conjugated pH/temperature/redox multi-stimuli responsive polymer microspheres for delivery of anti-cancer drug.

The folic acid (FA)-conjugated pH/temperature/redox multi-stimuli responsive poly(methacrylic acid-co-N,N-bis(acryloyl)cystamine/poly(N-isopropylacrylamide-co-glycidyl methacrylate-co-N,N-bis(acryloyl)cystamine) microspheres were prepared by a two-stage distillation-precipitation polymerization with subsequent surface modification with FA. The microspheres were characterized by transmission electron microscopy, dynamical light scattering, Fourier-transform infrared spectra, UV-vis spectra and elemental analysis. The degradation of the functional microspheres could be triggered by a reductive reagent, such as glutathione, due to presence of BAC crosslinker. The drug-loaded microspheres exhibited a pH/temperature/redox multi-stimuli responsive drug release character for doxorubicin hydrochloride as a model anti-cancer drug, which was efficiently loaded into the microspheres with a high loading capacity of 208.0% and an encapsulation efficiency of 85.4%. In vitro drug delivery study indicated that the FA-conjugated microspheres could deliver Dox into MCF-7 cells more efficiently than the microspheres without functionalization of FA. Furthermore, WST-1 assay showed that the microspheres had no obvious toxicity to MCF-7 cells even at a high concentration of 2000 µg mL(-1). The resultant microsphere may be a promising vector for delivery of anti-cancer drugs as it exhibits a low cytotoxicity and degradability, precise molecular targeting property and multi-stimuli responsively controlled drug release.