Human-multimodal deep learning collaboration in 'precise' diagnosis of lupus erythematosus subtypes and similar skin diseases

BACKGROUND

Lupus erythematosus (LE) is a spectrum of autoimmune diseases. Due to the complexity of cutaneous LE (CLE), clinical skin image-based artificial intelligence is still experiencing difficulties in distinguishing subtypes of LE.

OBJECTIVES

We aim to develop a multimodal deep learning system (MMDLS) for human-AI collaboration in diagnosis of LE subtypes.

METHODS

This is a multi-centre study based on 25 institutions across China to assist in diagnosis of LE subtypes, other eight similar skin diseases and healthy subjects. In total, 446 cases with 800 clinical skin images, 3786 multicolor-immunohistochemistry (multi-IHC) images and clinical data were collected, and EfficientNet-B3 and ResNet-18 were utilized in this study.

RESULTS

In the multi-classification task, the overall performance of MMDLS on 13 skin conditions is much higher than single or dual modals (Sen = 0.8288, Spe = 0.9852, Pre = 0.8518, AUC = 0.9844). Further, the MMDLS-based diagnostic-support help improves the accuracy of dermatologists from 66.88% ± 6.94% to 81.25% ± 4.23% (p = 0.0004).

CONCLUSIONS

These results highlight the benefit of human-MMDLS collaborated framework in telemedicine by assisting dermatologists and rheumatologists in the differential diagnosis of LE subtypes and similar skin diseases.

MA-PEP: A novel anticancer peptide prediction framework with multimodal feature fusion based on attention mechanism

AntiCancer Peptides (ACPs) have emerged as promising therapeutic agents for cancer treatment. The time-consuming and costly nature of wet-lab discriminatory methods has spurred the development of various machine learning and deep learning-based ACP classification methods. Nonetheless, current methods encountered challenges in efficiently integrating features from various peptide modalities, thereby limiting a more comprehensive understanding of ACPs and further restricting the improvement of prediction model performance. In this study, we introduce a novel ACP prediction method, MA-PEP, which leverages multiple attention mechanisms for feature enhancement and fusion to improve ACP prediction. By integrating the enhanced molecular-level chemical features and sequence information of peptides, MA-PEP demonstrates superior prediction performance across several benchmark datasets, highlighting its efficacy in ACP prediction. Moreover, the visual analysis and case studies further demonstrate MA-PEP's reliable feature extraction capability and its promise in the realm of ACP exploration. The code and datasets for MA-PEP are available at https://github.com/liangxiaodata/MA-PEP.

GraphPath: a graph attention model for molecular stratification with interpretability based on the pathway-pathway interaction network

MOTIVATION

Studying the molecular heterogeneity of cancer is essential for achieving personalized therapy. At the same time, understanding the biological processes that drive cancer development can lead to the identification of valuable therapeutic targets. Therefore, achieving accurate and interpretable clinical predictions requires paramount attention to thoroughly characterizing patients at both the molecular and biological pathway levels.

RESULTS

Here, we present GraphPath, a biological knowledge-driven graph neural network with multi-head self-attention mechanism that implements the pathway-pathway interaction network. We train GraphPath to classify the cancer status of patients with prostate cancer based on their multi-omics profiling. Experiment results show that our method outperforms P-NET and other baseline methods. Besides, two external cohorts are used to validate that the model can be generalized to unseen samples with adequate predictive performance. We reduce the dimensionality of latent pathway embeddings and visualize corresponding classes to further demonstrate the optimal performance of the model. Additionally, since GraphPath's predictions are interpretable, we identify target cancer-associated pathways that significantly contribute to the model's predictions. Such a robust and interpretable model has the potential to greatly enhance our understanding of cancer's biological mechanisms and accelerate the development of targeted therapies.

AVAILABILITY AND IMPLEMENTATION

https://github.com/amazingma/GraphPath.

Extracellular vesicles from UTX-knockout endothelial cells boost neural stem cell differentiation in spinal cord injury

BACKGROUND

Vascular endothelial cells are pivotal in the pathophysiological progression following spinal cord injury (SCI). The UTX (Ubiquitously Transcribed Tetratripeptide Repeat on Chromosome X) serves as a significant regulator of endothelial cell phenotype. The manipulation of endogenous neural stem cells (NSCs) offers a compelling strategy for the amelioration of SCI.

METHODS

Two mouse models were used to investigate SCI: NSCs lineage-traced mice and mice with conditional UTX knockout (UTX KO) in endothelial cells. To study the effects of UTX KO on neural differentiation, we harvested extracellular vesicles (EVs) from both UTX KO spinal cord microvascular endothelial cells (SCMECs) and negative control SCMECs. These EVs were then employed to modulate the differentiation trajectory of endogenous NSCs in the SCI model.

RESULTS

In our NSCs lineage-traced mice model of SCI, a marked decrease in neurogenesis was observed post-injury. Notably, NSCs in UTX KO SCMECs mice showed enhanced neuronal differentiation compared to controls. RNA sequencing and western blot analyses revealed an upregulation of L1 cell adhesion molecule (L1CAM), a gene associated with neurogenesis, in UTX KO SCMECs and their secreted EVs. This aligns with the observed promotion of neurogenesis in UTX KO conditions. In vivo administration of L1CAM-rich EVs from UTX KO SCMECs (KO EVs) to the mice significantly enhanced neural differentiation. Similarly, in vitro exposure of NSCs to KO EVs resulted in increased activation of the Akt signaling pathway, further promoting neural differentiation. Conversely, inhibiting Akt phosphorylation or knocking down L1CAM negated the beneficial effects of KO EVs on NSC neuronal differentiation.

CONCLUSIONS

In conclusion, our findings substantiate that EVs derived from UTX KO SCMECs can act as facilitators of neural differentiation following SCI. This study not only elucidates a novel mechanism but also opens new horizons for therapeutic interventions in the treatment of SCI. Video Abstract.

Inhibition of UTX/KDM6A improves recovery of spinal cord injury by attenuating BSCB permeability and macrophage infiltration through the MLCK/p-MLC pathway

Spinal cord injury (SCI) can prompt an immediate disruption to the blood-spinal cord barrier (BSCB). Restoring the integrity of this barrier is vital for the recovery of neurological function post-SCI. The UTX protein, a histone demethylase, has been shown in previous research to promote vascular regeneration and neurological recovery in mice with SCI. However, it is unclear whether UTX knockout could facilitate the recovery of the BSCB by reducing its permeability. In this study, we systematically studied BSCB disruption and permeability at different time points after SCI and found that conditional UTX deletion in endothelial cells (ECs) can reduce BSCB permeability, decrease inflammatory cell infiltration and ROS production, and improve neurological function recovery after SCI. Subsequently, we used RNA sequencing and ChIP-qPCR to confirm that conditional UTX knockout in ECs can down-regulate expression of myosin light chain kinase (MLCK), which specifically mediates myosin light chain (MLC) phosphorylation and is involved in actin contraction, cell retraction, and tight junctions (TJs) protein integrity. Moreover, we found that MLCK overexpression can increase the ratio of p-MLC/MLC, further break TJs, and exacerbate BSCB deterioration. Overall, our findings indicate that UTX knockout could inhibit the MLCK/p-MLC pathway, resulting in decreased BSCB permeability, and ultimately promoting neurological recovery in mice. These results suggest that UTX is a promising new target for treating SCI.

SSLpheno: a self-supervised learning approach for gene-phenotype association prediction using protein-protein interactions and gene ontology data

MOTIVATION

Medical genomics faces significant challenges in interpreting disease phenotype and genetic heterogeneity. Despite the establishment of standardized disease phenotype databases, computational methods for predicting gene-phenotype associations still suffer from imbalanced category distribution and a lack of labeled data in small categories.

RESULTS

To address the problem of labeled-data scarcity, we propose a self-supervised learning strategy for gene-phenotype association prediction, called SSLpheno. Our approach utilizes an attributed network that integrates protein-protein interactions and gene ontology data. We apply a Laplacian-based filter to ensure feature smoothness and use self-supervised training to optimize node feature representation. Specifically, we calculate the cosine similarity of feature vectors and select positive and negative sample nodes for reconstruction training labels. We employ a deep neural network for multi-label classification of phenotypes in the downstream task. Our experimental results demonstrate that SSLpheno outperforms state-of-the-art methods, especially in categories with fewer annotations. Moreover, our case studies illustrate the potential of SSLpheno as an effective prescreening tool for gene-phenotype association identification.

AVAILABILITY AND IMPLEMENTATION

https://github.com/bixuehua/SSLpheno.

UTX deletion promotes M2 macrophage polarization by epigenetically regulating endothelial cell-macrophage crosstalk after spinal cord injury

Macrophages polarized to the M2 subtype after spinal cord injury (SCI) are beneficial for promoting neurological recovery. The crosstalk between endothelial cells (ECs) and macrophages is crucial for the imbalance between proinflammatory and pro-resolving responses caused by macrophage heterogeneity; however, this crosstalk is strengthened post-SCI, leading to inflammatory cascades and second damage. As a powerful means to regulate gene expression, epigenetic regulation of the interaction between immune cells and ECs in SCI is still largely unknown. Our previous research demonstrated that the histone demethylase UTX deletion in ECs (UTX^-/- ECs) promotes neurological recovery, while the precise mechanism is unrevealed. Here, we discovered that UTX^-/- ECs polarize macrophages toward the M2 subtype post-SCI. Macrophage deficiency could block the neurological recovery caused by the knockdown of UTX. The exosomes from UTX^-/- ECs mediate this crosstalk. In addition, we found UTX, H3K27, and miR-467b-3p/Sfmbt2 promoters forming a regulatory complex that upregulates the miR-467b-3p in UTX^-/- ECs. And then, miR-467b-3p transfers to macrophages by exosomes and activates the PI3K/AKT/mTOR signaling by decreasing PTEN expression, finally polarizing macrophage to the M2 subtype. This study reveals a mechanism by epigenetic regulation of ECs-macrophages crosstalk and identifies potential targets, which may provide opportunities for treating SCI.