Deep generative modeling and clustering of single cell Hi-C data.

Deciphering 3D genome conformation is important for understanding gene regulation and cellular function at a spatial level. The recent advances of single cell Hi-C technologies have enabled the profiling of the 3D architecture of DNA within individual cell, which allows us to study the cell-to-cell variability of 3D chromatin organization. Computational approaches are in urgent need to comprehensively analyze the sparse and heterogeneous single cell Hi-C data. Here, we proposed scDEC-Hi-C, a new framework for single cell Hi-C analysis with deep generative neural networks. scDEC-Hi-C outperforms existing methods in terms of single cell Hi-C data clustering and imputation. Moreover, the generative power of scDEC-Hi-C could help unveil the differences of chromatin architecture across cell types. We expect that scDEC-Hi-C could shed light on deepening our understanding of the complex mechanism underlying the formation of chromatin contacts.

SAM-DTA: a sequence-agnostic model for drug-target binding affinity prediction.

Drug-target binding affinity prediction is a fundamental task for drug discovery and has been studied for decades. Most methods follow the canonical paradigm that processes the inputs of the protein (target) and the ligand (drug) separately and then combines them together. In this study we demonstrate, surprisingly, that a model is able to achieve even superior performance without access to any protein-sequence-related information. Instead, a protein is characterized completely by the ligands that it interacts. Specifically, we treat different proteins separately, which are jointly trained in a multi-head manner, so as to learn a robust and universal representation of ligands that is generalizable across proteins. Empirical evidences show that the novel paradigm outperforms its competitive sequence-based counterpart, with the Mean Squared Error (MSE) of 0.4261 versus 0.7612 and the R-Square of 0.7984 versus 0.6570 compared with DeepAffinity. We also investigate the transfer learning scenario where unseen proteins are encountered after the initial training, and the cross-dataset evaluation for prospective studies. The results reveals the robustness of the proposed model in generalizing to unseen proteins as well as in predicting future data. Source codes and data are available at https://github.com/huzqatpku/SAM-DTA.

Four and a half LIM domains 2 (FHL2) attenuates tumorigenesis of gastrointestinal stromal tumors (GISTs) by negatively regulating KIT signaling.

Gastrointestinal stromal tumors (GISTs) are predominately induced by KIT mutants. In this study, we found that four and a half LIM domains 2 (FHL2) was highly expressed in GISTs and KIT signaling dramatically increased FHL2 transcription while FHL2 inhibited KIT transcription. In addition, our results showed that FHL2 associated with KIT and increased the ubiquitination of both wild-type KIT and primary KIT mutants in GISTs, leading to decreased expression and activation of KIT although primary KIT mutants were less inhibited by FHL2 than wild-type KIT. In the animal experiments, loss of FHL2 expression in mice carrying germline KIT/V558A mutation which can develop GISTs resulted in increased tumor growth, but increased sensitivity of GISTs to imatinib treatment which is used as the first-line targeted therapy of GISTs, suggesting that FHL2 plays a role in the response of GISTs to KIT inhibitor. Unlike wild-type KIT and primary KIT mutants, we further found that FHL2 didn't alter the expression and activation of drug-resistant secondary KIT mutants. Taken together, our results indicated that FHL2 acts as the negative feedback of KIT signaling in GISTs while primary KIT mutants are less sensitive and secondary KIT mutants are resistant to the inhibition of FHL2.

Cell-permeable PI3 kinase competitive peptide inhibits KIT mutant mediated tumorigenesis of gastrointestinal stromal tumor (GIST).

BACKGROUND: Mutations in the receptor tyrosine kinase KIT are the main cause of gastrointestinal stromal tumor (GIST), and the KIT mutants mediated PI3 kinase activation plays a key role in the tumorigenesis of GIST. In this study, we aimed to block PI3 kinase activation by cell-permeable peptide and investigate its possible application in the treatment of GIST. METHODS AND RESULTS: We designed cell-permeable peptides based on the binding domain of PI3 kinase subunit p85 to KIT or PI3 kinase subunit p110, respectively, in order to compete for the binding between p85 and KIT or p110 and therefore inhibit the activation of PI3 kinases mediated by KIT. The results showed that the peptide can penetrate the cells, and inhibit the activation of PI3 kinases, leading to reduced cell survival and cell proliferation mediated by KIT mutants in vitro. Treatment of mice carrying germline KIT/V558A mutation, which can develop GIST, with the peptide that can compete for the binding between p85 and p110, led to reduced tumorigenesis of GIST. The peptide can further enhance the inhibition of the tumor growth by imatinib which is used as the first line targeted therapy of GIST. CONCLUSIONS: Our results showed that cell-permeable PI3 kinase competitive peptide can inhibit KIT-mediated PI3 kinase activation and tumorigenesis of GIST, providing a rationale to further test the peptide in the treatment of GIST and even other tumors with over-activation of PI3 kinases.

Prediction of tumor budding in patients with rectal adenocarcinoma using b-value threshold map.

OBJECTIVE: To investigate the feasibility of b-value threshold (bThreshold) map in preoperative evaluation of tumor budding (TB) in patients with locally advanced rectal cancer (LARC). METHODS: Patients with LARC were enrolled and underwent diffusion-weighted imaging (DWI). Contrast-to-noise ratio (CNR) between the lesions and normal tissues was assessed using DWI and bThreshold maps. TB was counted and scored using hematoxylin and eosin staining. Reproducibility for the apparent diffusion coefficient (ADC), bThreshold values, and region-of-interest (ROI) sizes were compared. Differences in ADC and bThreshold values with low-intermediate and high TB grades and the correlations between mean ADC and bThreshold values with TB categories were analyzed. Diagnostic performance of ADC and bThreshold values was assessed using area under the curve (AUC) and decision curve analysis. RESULTS: Fifty-one patients were evaluated. The CNR on bThreshold maps was significantly higher than that on DW images (9.807 ± 4.811 vs 7.779 ± 3.508, p = 0.005). Reproducibility was excellent for the ADC (ICC 0.933; CV 8.807%), bThreshold values (ICC 0.958; CV 7.399%), and ROI sizes (ICC 0.934; CV 8.425%). Significant negative correlations were observed between mean ADC values and TB grades and positive correlations were observed between mean bThreshold values and TB grades (p < 0.05). bThreshold maps showed better diagnostic performance than ADC maps (AUC, 0.914 vs 0.726; p = 0.048). CONCLUSIONS: In LARC patients, bThreshold values could distinguish different TB grades better than ADC values, and bThreshold maps may be a preoperative, non-invasive approach to evaluate TB grades. KEY POINTS: • Compared with diffusion-weighted images, bThreshold maps improved visualization and detection of rectal tumors. • Agreement and diagnostic performance of bThreshold values are superior to apparent diffusion coefficient in assessing tumor budding grades in patients with locally advanced rectal cancer. • bThreshold maps could be used to evaluate tumor budding grades non-invasively before operation.

SPRY4 inhibits and sensitizes the primary KIT mutants in gastrointestinal stromal tumors (GISTs) to imatinib.

BACKGROUND: KIT is frequently mutated in gastrointestinal stromal tumors (GISTs), and the treatment of GISTs largely relies on targeting KIT currently. In this study, we aimed to investigate the role of sprouty RTK signaling antagonist 4 (SPRY4) in GISTs and related mechanisms. METHODS: Ba/F3 cells and GIST-T1 cell were used as cell models, and mice carrying germline KIT/V558A mutation were used as animal model. Gene expression was examined by qRT-PCR and western blot. Protein association was examined by immunoprecipitation. RESULTS: Our study revealed that KIT increased the expression of SPRY4 in GISTs. SPRY4 was found to bind to both wild-type KIT and primary KIT mutants in GISTs, and inhibited KIT expression and activation, leading to decreased cell survival and proliferation mediated by KIT. We also observed that inhibition of SPRY4 expression in KITV558A/WT mice led to increased tumorigenesis of GISTs in vivo. Moreover, our results demonstrated that SPRY4 enhanced the inhibitory effect of imatinib on the activation of primary KIT mutants, as well as on cell proliferation and survival mediated by the primary KIT mutants. However, in contrast to this, SPRY4 did not affect the expression and activation of drug-resistant secondary KIT mutants, nor did it affect the sensitivity of secondary KIT mutants to imatinib. These findings suggested that secondary KIT mutants regulate a different downstream signaling cascade than primary KIT mutants. CONCLUSIONS: Our results suggested that SPRY4 acts as negative feedback of primary KIT mutants in GISTs by inhibiting KIT expression and activation. It can increase the sensitivity of primary KIT mutants to imatinib. In contrast, secondary KIT mutants are resistant to the inhibition of SPRY4.

Predictive value of modified MRI-based split scar sign (mrSSS) score for pathological complete response after neoadjuvant chemoradiotherapy for patients with rectal cancer.

PURPOSE: To measure the diagnostic performance of modified MRI-based split scar sign (mrSSS) score for the prediction of pathological complete response (pCR) after neoadjuvant chemoradiotherapy (nCRT) for patients with rectal cancer. METHODS: The modified MRI-based split scar sign (mrSSS) score, which consists of T2-weighted images (T2WI)-based score and diffusion-weighted images (DWI)-based score. The sensitivity, specificity, and accuracy of modified mrSSS score, endoscopic gross type, and MRI-based tumor regression grading (mrTRG) score, in the prediction of pCR, were compared. The prognostic value of the modified mrSSS score was also studied. RESULTS: A total of 189 patients were included in the study. The Kendall's coefficient of interobserver concordance of modified mrSSS score, T2WI -based score, and DWI-based score were 0.899, 0.890, and 0.789 respectively. And the maximum and minimum k value of the modified mrSSS score was 0.797 (0.742-0.853) and 0.562 (0.490-0.634). The sensitivity, specificity, and accuracy of prediction of pCR were 0.66, 0.97, and 0.90 for modified mrSSS score; 0.37, 0.89, and 0.78 for endoscopic gross type (scar); and 0.24, 0.92, and 0.77 for mrTRG score (mrTRG = 1). The modified mrSSS score had significantly higher sensitivity than the endoscopic gross type and the mrTRG score in predicting pCR. Patients with lower modified mrSSS scores had significantly longer disease-free survival (P < 0.05). CONCLUSION: The modified mrSSS score showed satisfactory interobserver agreement and higher sensitivity in predicting pCR after nCRT in patients with rectal cancer. The modified mrSSS score is also a predictor of disease-free survival.

TISS-net: Brain tumor image synthesis and segmentation using cascaded dual-task networks and error-prediction consistency.

Accurate segmentation of brain tumors from medical images is important for diagnosis and treatment planning, and it often requires multi-modal or contrast-enhanced images. However, in practice some modalities of a patient may be absent. Synthesizing the missing modality has a potential for filling this gap and achieving high segmentation performance. Existing methods often treat the synthesis and segmentation tasks separately or consider them jointly but without effective regularization of the complex joint model, leading to limited performance. We propose a novel brain Tumor Image Synthesis and Segmentation network (TISS-Net) that obtains the synthesized target modality and segmentation of brain tumors end-to-end with high performance. First, we propose a dual-task-regularized generator that simultaneously obtains a synthesized target modality and a coarse segmentation, which leverages a tumor-aware synthesis loss with perceptibility regularization to minimize the high-level semantic domain gap between synthesized and real target modalities. Based on the synthesized image and the coarse segmentation, we further propose a dual-task segmentor that predicts a refined segmentation and error in the coarse segmentation simultaneously, where a consistency between these two predictions is introduced for regularization. Our TISS-Net was validated with two applications: synthesizing FLAIR images for whole glioma segmentation, and synthesizing contrast-enhanced T1 images for Vestibular Schwannoma segmentation. Experimental results showed that our TISS-Net largely improved the segmentation accuracy compared with direct segmentation from the available modalities, and it outperformed state-of-the-art image synthesis-based segmentation methods.

MRI-based nomogram analysis: recognition of anterior peritoneal reflection and its relationship to rectal cancers.

BACKGROUND: This study is aimed to explore the factors influencing the visualization of the anterior peritoneal reflection (APR) and evaluated the feasibility of measuring the distance from the anal verge to APR (AV-APR), the tumor height on MRI and the accuracy of determining the tumor location with regard to APR. METHODS: We retrospectively analyzed 110 patients with rectal cancer. A univariate and multivariate logistic regression was performed to identify the independent factors (age, sex, T stage, the degree of bladder filling, pelvic effusion, intraoperative tumor location, BMI, uterine orientation, the distance from seminal vesicle/uterus to rectum) associated with the visualization of the APR on MRI. The nomogram diagram and receiver operating characteristic curve (ROC curve) were established. Intraclass correlation coefficient (ICC) was used to evaluate the consistency of the distance of AV-APR. The Pearson correlation coefficient was used to characterize the agreement between measurements of the tumor height by colonoscopy and MRI. The Kappa statistics was used to evaluate the value of MRI in the diagnosis of the tumor location with regard to the APR. RESULTS: Multivariate logistic regression showed that BMI (P = 0.031, odds ratio, OR = 1.197), pelvic effusion (P = 0.020, OR = 7.107) and the distance from seminal vesicle/uterus to the rectum (P = 0.001, OR = 3.622) were correlated with the visualization of APR. The cut-off point of BMI and the distance from seminal vesicle/uterus to the rectum is 25.845 kg/m2 and 1.15 cm. The area under curve (AUC) (95% Confidence Interval, 95% CI) of the combined model is 0.840 (0.750-0.930). The favorable calibration of the nomogram showed a non-significant Hosmer-Lemeshow test statistic (P = 0.195). The ICC value (95% CI) of the distance of AV-APR measured by two radiologists was 0.981 (0.969-0.989). The height measured by MRI and colonoscopy were correlated with each other (r = 0.699, P < 0.001). The Kappa value was 0.854. CONCLUSIONS: BMI, pelvic effusion, and the distance from seminal vesicle/uterus to rectum could affect the visualization of APR on MRI. Also, it's feasible to measure the distance of AV-APR, the tumor height, and to evaluate the tumor location with regard to APR using MRI.

Multiparametric Magnetic Resonance Imaging Features That Predict Treatment Response to Endovascular Sclerotherapy in Craniofacial Venous Malformations.

PURPOSE: Craniofacial venous malformations (VMs) cause severe psychological and physiological burden to patients, and treatment is meaningful only when the benefits of treatment outweigh the risks. Therefore, it is very important to predict the treatment response before treatment. This study was performed to explore the value of multiparametric magnetic resonance imaging for predicting treatment response to endovascular sclerotherapy in VMs. MATERIALS AND METHODS: We designed and implemented a case-control study and enrolled a sample from patients with VM treated by endovascular sclerotherapy at our hospital from January 2014 to January 2018. The primary predictor variables were pretreatment volume (prevolume), lesion classification, phleboliths, initial slope of increase (ISI), gender, age, and sclerosants. The primary outcome variable was treatment response (positive response or negative response). Descriptive, univariate and multivariate binary logistic regressions, and Firth's penalized maximum likelihood estimate were computed to measure the association between predictor variables and treatment response. The level of statistical significance was set at a P value less than or equal to .05. RESULTS: The sample was composed of 42 patients with a median age of 17.50 years, and 33.3% were males. There were 27 and 15 patients in the positive and negative response groups, respectively. There were significant differences between the 2 groups for ISI (adjusted odds ratio [OR], 2.184; P = .0268; 95% confidence interval [95% CI], 1.094 to 4.360), lesion classification (adjusted OR, 9.072; P = .0226; 95% CI, 1.363 to 60.400), and prevolume (adjusted OR, 1.020; P = .0268; 95% CI, 1.002 to 1.038). The cutoff point for prevolume and ISI was 40.42 cm3 and 2.61. CONCLUSIONS: Multiparametric magnetic resonance imaging could provide an approach for predicting treatment response in craniofacial VMs. When the prevolume was greater than 40.42 cm3, ISI was greater than 2.61, and the classification was infiltrating type, the response to sclerotherapy was negative.

Computational modeling of cellular structures using conditional deep generative networks.

MOTIVATION: Cellular function is closely related to the localizations of its sub-structures. It is, however, challenging to experimentally label all sub-cellular structures simultaneously in the same cell. This raises the need of building a computational model to learn the relationships among these sub-cellular structures and use reference structures to infer the localizations of other structures. RESULTS: We formulate such a task as a conditional image generation problem and propose to use conditional generative adversarial networks for tackling it. We employ an encoder-decoder network as the generator and propose to use skip connections between the encoder and decoder to provide spatial information to the decoder. To incorporate the conditional information in a variety of different ways, we develop three different types of skip connections, known as the self-gated connection, encoder-gated connection and label-gated connection. The proposed skip connections are built based on the conditional information using gating mechanisms. By learning a gating function, the network is able to control what information should be passed through the skip connections from the encoder to the decoder. Since the gate parameters are also learned automatically, we expect that only useful spatial information is transmitted to the decoder to help image generation. We perform both qualitative and quantitative evaluations to assess the effectiveness of our proposed approaches. Experimental results show that our cGAN-based approaches have the ability to generate the desired sub-cellular structures correctly. Our results also demonstrate that the proposed approaches outperform the existing approach based on adversarial auto-encoders, and the new skip connections lead to improved performance. In addition, the localizations of generated sub-cellular structures by our approaches are consistent with observations in biological experiments. AVAILABILITY AND IMPLEMENTATION: The source code and more results are available at https://github.com/divelab/cgan/.

Interactive Exploration for Continuously Expanding Neuron Databases.

This paper proposes a novel framework to help biologists explore and analyze neurons based on retrieval of data from neuron morphological databases. In recent years, the continuously expanding neuron databases provide a rich source of information to associate neuronal morphologies with their functional properties. We design a coarse-to-fine framework for efficient and effective data retrieval from large-scale neuron databases. In the coarse-level, for efficiency in large-scale, we employ a binary coding method to compress morphological features into binary codes of tens of bits. Short binary codes allow for real-time similarity searching in Hamming space. Because the neuron databases are continuously expanding, it is inefficient to re-train the binary coding model from scratch when adding new neurons. To solve this problem, we extend binary coding with online updating schemes, which only considers the newly added neurons and update the model on-the-fly, without accessing the whole neuron databases. In the fine-grained level, we introduce domain experts/users in the framework, which can give relevance feedback for the binary coding based retrieval results. This interactive strategy can improve the retrieval performance through re-ranking the above coarse results, where we design a new similarity measure and take the feedback into account. Our framework is validated on more than 17,000 neuron cells, showing promising retrieval accuracy and efficiency. Moreover, we demonstrate its use case in assisting biologists to identify and explore unknown neurons.

Groupwise registration of sequential images from multispectral imaging (MSI) of the retina and choroid.

Multispectral Imaging (MSI) produces a sequence of discrete spectral slices that penetrate different light-absorbing species or chromophores and is a noninvasive technology useful for the early detection of various retinal, optic nerve and choroidal diseases. However, eye movement during the image acquisition process may introduce spatial misalignment between MSI images. This potentially causes trouble in the manual/automatic interpretation of MSI, but still remains an unresolved problem to this date. To deal with this MSI misalignment problem, we present a method on the groupwise registration of sequential images from MSI of the retina and choroid. The advantage of our algorithm is at least threefold: 1) simultaneous estimation of landmark correspondences and a parametric motion model via quadratic programming, 2) enforcement of temporal smoothness on the estimated motion, and 3) inclusion of a robust matching cost function. As validated in our experiments with a database of 22 MSI sequences, our algorithm outperforms two state-of-the-art registration techniques proposed originally in other domains. Our algorithm is potentially invaluable in ophthalmologists' clinical practice regarding various eye diseases.

On the challenges and perspectives of foundation models for medical image analysis.

This article discusses the opportunities, applications and future directions of large-scale pretrained models, i.e., foundation models, which promise to significantly improve the analysis of medical images. Medical foundation models have immense potential in solving a wide range of downstream tasks, as they can help to accelerate the development of accurate and robust models, reduce the dependence on large amounts of labeled data, preserve the privacy and confidentiality of patient data. Specifically, we illustrate the "spectrum" of medical foundation models, ranging from general imaging models, modality-specific models, to organ/task-specific models, and highlight their challenges, opportunities and applications. We also discuss how foundation models can be leveraged in downstream medical tasks to enhance the accuracy and efficiency of medical image analysis, leading to more precise diagnosis and treatment decisions.

Pattern-Aware Transformer: Hierarchical Pattern Propagation in Sequential Medical Images.

This paper investigates how to effectively mine contextual information among sequential images and jointly model them in medical imaging tasks. Different from state-of-the-art methods that model sequential correlations via point-wise token encoding, this paper develops a novel hierarchical pattern-aware tokenization strategy. It handles distinct visual patterns independently and hierarchically, which not only ensures the full flexibility of attention aggregation under different pattern representations but also preserves both local and global information simultaneously. Based on this strategy, we propose a Pattern-Aware Transformer (PATrans) featuring a global-local dual-path pattern-aware cross-attention mechanism to achieve hierarchical pattern matching and propagation among sequential images. Furthermore, PATrans is plug-and-play and can be seamlessly integrated into various backbone networks for diverse downstream sequence modeling tasks. We demonstrate its general application paradigm across four domains and five benchmarks in video object detection and 3D volumetric semantic segmentation tasks, respectively. Impressively, PATrans sets new state-of-the-art across all these benchmarks, i.e., CVC-Video (92.3% detection F1), ASU-Mayo (99.1% localization F1), Lung Tumor (78.59% DSC), Nasopharynx Tumor (75.50% DSC), and Kidney Tumor (87.53% DSC). Codes and models are available at https://github.com/GGaoxiang/PATrans.

Development of a CT-assessed adiposity nomogram for predicting outcome in localized ccRCC.

PURPOSE: This study aimed to develop and validate a computed tomography-based nomogram assessing visceral and subcutaneous adiposity for predicting outcomes in localized clear cell renal cell carcinoma (ccRCC). METHODS: A cohort of 364 patients with pathologically confirmed ccRCC participated in this retrospective study, with 254 patients assigned to the training set and 110 to the validation set (a 7:3 distribution ratio). The adipose score (AS) was generated using the least absolute shrinkage and selection operator Cox regression. Subsequently, a nomogram was constructed by integrating the clinical independent predictor with the AS to predict disease-free survival (DFS) in localized ccRCC after surgery. The performance of the nomogram was compared with the University of California, Los Angeles, Integrated Staging System (UISS), and the Stage, Size, Grade, and Necrosis (SSIGN) score. RESULTS: In both the training and validation cohorts, the nomogram exhibited superior discrimination compared to SSIGN and UISS (C-index: 0.897 vs. 0.781 vs. 0.776 in the training cohort, and 0.752 vs. 0.596 vs. 0.686 in the validation cohort; 5 year AUC: 0.907 vs. 0.805 vs. 0.820 in the training cohort, and 0.832 vs. 0.577 vs. 0.726 in the validation cohort). Decision curve analysis (DCA) revealed a superior net benefit across a wider range of threshold probabilities for predicting 5 year DFS compared to UISS and SSIGN scores. CONCLUSIONS: The developed prognostic nomogram demonstrated high accuracy and overall superior performance compared to existing prognostic models.

Synthesis of C-N Axially Chiral N-Arylbenzo[g]indoles via a Central-to-Axial Chirality Conversion Strategy.

Gold-catalyzed cascade cyclization of diynes for the synthesis of previously unexplored C-N axially chiral N-arylbenzo[g]indoles was described. The transformation was achieved via a central-to-axial chirality conversion strategy. The chiral conversion exhibited high efficiency. Besides single C-N chiral axis, N-arylbenzo[g]indoles bearing both C-N and C-C chiral axes were also afforded. The title compound derived monophosphine ligand was prepared and was evaluated in Pd-catalyzed asymmetric allylic substitutions, showing excellent chiral induction ability.

Hierarchical-Instance Contrastive Learning for Minority Detection on Imbalanced Medical Datasets.

Deep learning methods are often hampered by issues such as data imbalance and data-hungry. In medical imaging, malignant or rare diseases are frequently of minority classes in the dataset, featured by diversified distribution. Besides that, insufficient labels and unseen cases also present conundrums for training on the minority classes. To confront the stated problems, we propose a novel Hierarchical-instance Contrastive Learning (HCLe) method for minority detection by only involving data from the majority class in the training stage. To tackle inconsistent intra-class distribution in majority classes, our method introduces two branches, where the first branch employs an auto-encoder network augmented with three constraint functions to effectively extract image-level features, and the second branch designs a novel contrastive learning network by taking into account the consistency of features among hierarchical samples from majority classes. The proposed method is further refined with a diverse mini-batch strategy, enabling the identification of minority classes under multiple conditions. Extensive experiments have been conducted to evaluate the proposed method on three datasets of different diseases and modalities. The experimental results show that the proposed method outperforms the state-of-the-art methods.

TMO-Net: an explainable pretrained multi-omics model for multi-task learning in oncology.

Cancer is a complex disease composing systemic alterations in multiple scales. In this study, we develop the Tumor Multi-Omics pre-trained Network (TMO-Net) that integrates multi-omics pan-cancer datasets for model pre-training, facilitating cross-omics interactions and enabling joint representation learning and incomplete omics inference. This model enhances multi-omics sample representation and empowers various downstream oncology tasks with incomplete multi-omics datasets. By employing interpretable learning, we characterize the contributions of distinct omics features to clinical outcomes. The TMO-Net model serves as a versatile framework for cross-modal multi-omics learning in oncology, paving the way for tumor omics-specific foundation models.

Classification of lung cancer subtypes on CT images with synthetic pathological priors.

The accurate diagnosis on pathological subtypes for lung cancer is of significant importance for the follow-up treatments and prognosis managements. In this paper, we propose self-generating hybrid feature network (SGHF-Net) for accurately classifying lung cancer subtypes on computed tomography (CT) images. Inspired by studies stating that cross-scale associations exist in the image patterns between the same case's CT images and its pathological images, we innovatively developed a pathological feature synthetic module (PFSM), which quantitatively maps cross-modality associations through deep neural networks, to derive the "gold standard" information contained in the corresponding pathological images from CT images. Additionally, we designed a radiological feature extraction module (RFEM) to directly acquire CT image information and integrated it with the pathological priors under an effective feature fusion framework, enabling the entire classification model to generate more indicative and specific pathologically related features and eventually output more accurate predictions. The superiority of the proposed model lies in its ability to self-generate hybrid features that contain multi-modality image information based on a single-modality input. To evaluate the effectiveness, adaptability, and generalization ability of our model, we performed extensive experiments on a large-scale multi-center dataset (i.e., 829 cases from three hospitals) to compare our model and a series of state-of-the-art (SOTA) classification models. The experimental results demonstrated the superiority of our model for lung cancer subtypes classification with significant accuracy improvements in terms of accuracy (ACC), area under the curve (AUC), positive predictive value (PPV) and F1-score.