The Innovation最新文献

Accurate profiling of microscopy images from small scale to high throughput is an essential procedure in basic and applied biological research. Here, we present Microsnoop, a novel deep learning–based representation tool trained on large-scale microscopy images using masked self-supervised learning. Microsnoop can process various complex and heterogeneous images, and we classified images into three categories: single-cell, full-field, and batch-experiment images. Our benchmark study on 10 high-quality evaluation datasets, containing over 2,230,000 images, demonstrated Microsnoop’s robust and state-of-the-art microscopy image representation ability, surpassing existing generalist and even several custom algorithms. Microsnoop can be integrated with other pipelines to perform tasks such as superresolution histopathology image and multimodal analysis. Furthermore, Microsnoop can be adapted to various hardware and can be easily deployed on local or cloud computing platforms. We will regularly retrain and reevaluate the model using community-contributed data to consistently improve Microsnoop.

Amyloid-β, tau pathology, and biomarkers of neurodegeneration make up the core diagnostic biomarkers of Alzheimer disease (AD). However, these proteins represent only a fraction of the complex biological processes underlying AD, and individuals with other brain diseases in which AD pathology is a comorbidity also test positive for these diagnostic biomarkers. More AD-specific early diagnostic and disease staging biomarkers are needed. In this study, we performed tandem mass tag proteomic analysis of paired cerebrospinal fluid (CSF) and serum samples in a discovery cohort comprising 98 participants. Candidate biomarkers were validated by parallel reaction monitoring–based targeted proteomic assays in an independent multicenter cohort comprising 288 participants. We quantified 3,238 CSF and 1,702 serum proteins in the discovery cohort, identifying 171 and 860 CSF proteins and 37 and 323 serum proteins as potential early diagnostic and staging biomarkers, respectively. In the validation cohort, 58 and 21 CSF proteins, as well as 12 and 18 serum proteins, were verified as early diagnostic and staging biomarkers, respectively. Separate 19-protein CSF and an 8-protein serum biomarker panels were built by machine learning to accurately classify mild cognitive impairment (MCI) due to AD from normal cognition with areas under the curve of 0.984 and 0.881, respectively. The 19-protein CSF biomarker panel also effectively discriminated patients with MCI due to AD from patients with other neurodegenerative diseases. Moreover, we identified 21 CSF and 18 serum stage-associated proteins reflecting AD stages. Our findings provide a foundation for developing blood-based tests for AD screening and staging in clinical practice.

Prostate cancer detected by autopsy is named latent prostate cancer. As the repertoire of clinical prostate cancer, latent cancer may better reflect the disease burden. Unlike clinical prostate specimens, which are obtained exclusively from biopsy-positive cases, prostate specimens obtained through autopsy provide information on biopsy-negative cases, helping calculate the true sensitivity of prostate biopsy. From 2014 to 2021, we collected autopsy specimens of the prostate from body donors in China and performed transperineal and transrectal biopsies on specimens before step-sectioning and pathological measurements. We found that the crude prevalence of latent prostate cancer in middle-aged and elderly men was 35.1% (81/231), which was higher than previous estimates for Chinese populations. The overall per-patient sensitivities of transperineal and transrectal biopsies were not significantly different (33.3% vs. 32.1%, p = 0.82), but the two approaches differed in preferential sampling area along the proximal-distal axis of the prostate. Transperineal biopsy had a higher sensitivity for detecting clinically significant lesions in the distal third (34.7% vs. 16.3%, p = 0.02) and distal half (30.6% vs. 18.1%, p = 0.04), while transrectal biopsy had a higher sensitivity for lesions in the proximal half (25.0% vs. 13.9%, p = 0.046). Both transperineal and transrectal methods of biopsy missed most small lesions (<0.1 mL) and 35.3% (6/17) of large lesions (>0.5 mL). In conclusion, the prevalence of latent prostate cancer in China has increased over the past 2 decades. Systematic transperineal and transrectal methods of biopsy had comparable sensitivities but had different preferential sampling areas. Both approaches miss one-third of large lesions.

The remote forcing from land surface changes in the Sahara is hypothesized to play a pivotal role in modulating the intensity of the East Asian summer monsoon (EASM) through ocean-atmospheric teleconnections. This modulation has far-reaching consequences, particularly in facilitating societal shifts documented in northern China. Here, we present a well-dated lake-level record from the Daihai Lake basin in northern China, providing quantitative assessments of Holocene monsoonal precipitation and the consequent migrations of the northern boundary of the EASM. Our reconstruction, informed by a water and energy balance model, indicates that annual precipitation reached ∼700 mm during 8-5 ka, followed by a rapid decline to ∼550 mm between 5-4 ka. This shift coherently aligns with a significant ∼300 km northwestward movement of the EASM northern boundary during the middle Holocene (MH), in contrast to its current position. Our findings underscore that these changes cannot be entirely attributed to orbital forcing, as corroborated by simulation tests. Climate model simulations deployed in our study suggest that the presence of the Green Sahara during the MH significantly strengthened the EASM and led to a northward shift of the monsoon rainfall belt. Conversely, the Sahara’s reversion to a desert landscape in the late Holocene was accompanied by a corresponding southward retraction of monsoon influence. These dramatic hydroclimate changes around 5-4 ka likely triggered or at least contributed to a shift in Neolithic cultures and societal transformation in northern China. With decreasing agricultural productivity, communities transitioned from millet farming to a mixed rain-fed agriculture and animal husbandry system. Thus, our findings elucidate not only the variability of the EASM but also the profound implications of a remote forcing, such as surface transformations of the Sahara, on climatic changes and cultural evolution in northern China.

Tightly integrating actuation, computation, and sensing in soft materials allows soft robots to respond autonomously to their environments. However, fusing these capabilities within a single soft module in an efficient, programmable, and compatible way is still a significant challenge. Herein, we introduce a strategy for integrating actuation, computation, and sensing capabilities in soft origami. Unified and plug-and-play soft origami modules can be reconfigured into diverse morphologies with specific functions or reprogrammed into a variety of soft logic circuits, similar to “Lego” bricks. We then built an untethered autonomous soft turtle that is able to sense stimuli, store data, process information, and perform swimming movements. The function multiplexing and signal compatibility of the origami minimize the number of soft devices, thereby reducing the complexity and redundancy of soft robots. Moreover, this origami also exhibits strong damage-resistance and high durability. We envision that this work will offer an effective way to readily create on-demand soft robots that can operate in unknown environments.