Background: Accurate and robust pathological image analysis for colorectal cancer (CRC) diagnosis is time-consuming and knowledge-intensive, but is essential for CRC patients’ treatment. The current heavy workload of pathologists in clinics/hospitals may easily lead to unconscious misdiagnosis of CRC based on daily image analyses. Methods: Based on a state-of-the-art transfer-learned deep convolutional neural network in artificial intelligence (AI), we proposed a novel patch aggregation strategy for clinic CRC diagnosis using weakly labeled pathological whole-slide image (WSI) patches. This approach was trained and validated using an unprecedented and enormously large number of 170,099 patches, > 14,680 WSIs, from > 9631 subjects that covered diverse and representative clinical cases from multi-independent-sources across China, the USA, and Germany. Results: Our innovative AI tool consistently and nearly perfectly agreed with (average Kappa statistic 0.896) and even often better than most of the experienced expert pathologists when tested in diagnosing CRC WSIs from multicenters. The average area under the receiver operating characteristics curve (AUC) of AI was greater than that of the pathologists (0.988 vs 0.970) and achieved the best performance among the application of other AI methods to CRC diagnosis. Our AI-generated heatmap highlights the image regions of cancer tissue/cells. Conclusions: This first-ever generalizable AI system can handle large amounts of WSIs consistently and robustly without potential bias due to fatigue commonly experienced by clinical pathologists. It will drastically alleviate the heavy clinical burden of daily pathology diagnosis and improve the treatment for CRC patients. This tool is generalizable to other cancer diagnosis based on image recognition.

Background: Increasing evidence suggests that human gut microbiome plays an important role in variation of skeletal muscle mass (SMM). However, specific causal mechanistic relationship of human gut microbiome with SMM remains largely unresolved. Understanding the causal mechanistic relationship may provide a basis for novel interventions for loss of SMM. This study investigated whether human gut microbiome has a causal effect on SMM among Chinese community‐dwelling healthy menopausal women.Methods: Estimated SMM was derived from whole‐body dual‐energy X‐ray absorptiometry. We performed integrated analyses on whole‐genome sequencing, shotgun metagenomic sequencing, and serum short‐chain fatty acids (SCFAs), as well as available host SMM measurements among community‐dwelling healthy menopausal women (N = 482). We combined the results with summary statistics from genome‐wide association analyses for human gut microbiome (N = 952) and SMM traits (N = 28 330). As a prerequisite for causality, we used a computational protocol that was proposed to measure correlations among gut metagenome, metabolome, and the host trait to investigate the relationship between human gut microbiome and SMM. Causal inference methods were applied to assess the potential causal effects of gut microbial features on SMM, through one‐sample and two‐sample Mendelian randomization (MR) analyses, respectively.Results: In metagenomic association analyses, the increased capacity for gut microbial synthesis of the SCFA butyrate was significantly associated with serum butyrate levels [Spearman correlation coefficient (SCC) = 0.13, P = 0.02] and skeletal muscle index (SCC = 0.084, P = 0.002). Of interest was the finding that two main butyrate‐producing bacterial species were both positively associated with the increased capacity for gut microbial synthesis of butyrate [Faecalibacterium prausnitzii (SCC = 0.25, P = 6.6 × 10⁻⁷) and Butyricimonas virosa (SCC = 0.15, P = 0.001)] and for skeletal muscle index [F. prausnitzii (SCC = 0.16, P = 6.2 × 10⁻⁴) and B. virosa (SCC = 0.17, P = 2.4 × 10⁻⁴)]. One‐sample MR results showed a causal effect between gut microbial synthesis of the SCFA butyrate and appendicular lean mass (β = 0.04, 95% confidence interval 0.029 to 0.051, P = 0.003). Two‐sample MR results further confirmed the causal effect between gut microbial synthesis of the SCFA butyrate and appendicular lean mass (β = 0.06, 95% confidence interval 0 to 0.13, P = 0.06).Conclusions: Our results may help the future development of novel intervention approaches for preventing or alleviating loss of SMM.

Machine-assisted pathological recognition has been focused on supervised learning (SL) that suffers from a significant annotation bottleneck. We propose a semi-supervised learning (SSL) method based on the mean teacher architecture using 13,111 whole slide images of colorectal cancer from 8803 subjects from 13 independent centers. SSL (~3150 labeled, ~40,950 unlabeled; ~6300 labeled, ~37,800 unlabeled patches) performs significantly better than the SL. No significant difference is found between SSL (~6300 labeled, ~37,800 unlabeled) and SL (~44,100 labeled) at patch-level diagnoses (area under the curve (AUC): 0.980 ± 0.014 vs. 0.987 ± 0.008, P value = 0.134) and patient-level diagnoses (AUC: 0.974 ± 0.013 vs. 0.980 ± 0.010, P value = 0.117), which is close to human pathologists (average AUC: 0.969). The evaluation on 15,000 lung and 294,912 lymph node images also confirm SSL can achieve similar performance as that of SL with massive annotations. SSL dramatically reduces the annotations, which has great potential to effectively build expert-level pathological artificial intelligence platforms in practice.

Bone marrow-derived mesenchymal stem cells (BM-MSCs) are multipotent stromal cells that have a critical role in the maintenance of skeletal tissues such as bone, cartilage, and the fat in bone marrow. In addition to providing microenvironmental support for hematopoietic processes, BM-MSCs can differentiate into various mesodermal lineages including osteoblast/osteocyte, chondrocyte, and adipocyte that are crucial for bone metabolism. While BM-MSCs have high cell-to-cell heterogeneity in gene expression, the cell subtypes that contribute to this heterogeneity in vivo in humans have not been characterized. To investigate the transcriptional diversity of BM-MSCs, we applied single-cell RNA sequencing (scRNA-seq) on freshly isolated CD271⁺ BM-derived mononuclear cells (BM-MNCs) from two human subjects. We successfully identified LEPR^hiCD45^low BM-MSCs within the CD271⁺ BM-MNC population, and further codified the BM-MSCs into distinct subpopulations corresponding to the osteogenic, chondrogenic, and adipogenic differentiation trajectories, as well as terminal-stage quiescent cells. Biological functional annotations of the transcriptomes suggest that osteoblast precursors induce angiogenesis coupled with osteogenesis, and chondrocyte precursors have the potential to differentiate into myocytes. We also discovered transcripts for several clusters of differentiation (CD) markers that were either highly expressed (e.g., CD167b, CD91, CD130 and CD118) or absent (e.g., CD74, CD217, CD148 and CD68) in BM-MSCs, representing potential novel markers for human BM-MSC purification. This study is the first systematic in vivo dissection of human BM-MSCs cell subtypes at the single-cell resolution, revealing an insight into the extent of their cellular heterogeneity and roles in maintaining bone homeostasis.

Although the gut microbiota has been reported to influence osteoporosis risk, the individual species involved, and underlying mechanisms, remain largely unknown. We performed integrative analyses in a Chinese cohort of peri-/post-menopausal women with metagenomics/targeted metabolomics/whole-genome sequencing to identify novel microbiome-related biomarkers for bone health. Bacteroides vulgatus was found to be negatively associated with bone mineral density (BMD), which was validated in US white people. Serum valeric acid (VA), a microbiota derived metabolite, was positively associated with BMD and causally downregulated by B. vulgatus. Ovariectomized mice fed B. vulgatus demonstrated increased bone resorption and poorer bone micro-structure, while those fed VA demonstrated reduced bone resorption and better bone micro-structure. VA suppressed RELA protein production (pro-inflammatory), and enhanced IL10 mRNA expression (anti-inflammatory), leading to suppressed maturation of osteoclast-like cells and enhanced maturation of osteoblasts in vitro. The findings suggest that B. vulgatus and VA may represent promising targets for osteoporosis prevention/treatment.