TUnA: an uncertainty-aware transformer model for sequence-based protein-protein interaction prediction.

Protein-protein interactions (PPIs) are important for many biological processes, but predicting them from sequence data remains challenging. Existing deep learning models often cannot generalize to proteins not present in the training set and do not provide uncertainty estimates for their predictions. To address these limitations, we present TUnA, a Transformer-based uncertainty-aware model for PPI prediction. TUnA uses ESM-2 embeddings with Transformer encoders and incorporates a Spectral-normalized Neural Gaussian Process. TUnA achieves state-of-the-art performance and, importantly, evaluates uncertainty for unseen sequences. We demonstrate that TUnA's uncertainty estimates can effectively identify the most reliable predictions, significantly reducing false positives. This capability is crucial in bridging the gap between computational predictions and experimental validation.

Highly porous hydrogels for efficient solar water evaporation.

Solar energy is a plentiful renewable resource on Earth, with versatile applications in both domestic and industrial settings, particularly in solar steam generation (SSG). However, current SSG processes encounter challenges such as low efficiency and the requirement for extremely high concentrations of solar irradiation. Interfacial evaporation technology has emerged as a solution to these issues, offering improved solar performance compared to conventional SSG processes. Nonetheless, its implementation introduces additional complexities and costs to system construction. In this study, we present the development of hydrophilic, three-dimensional network-structured hydrogels with high porosity and swelling ratio using a facile fabrication technique. We systematically varied the mixing ratios of four key ingredients (polyethylene glycol diacrylate, PEGDA; polyethylene glycol methyl-ether acrylate, PEGMA; phosphate-buffered saline, PBS; and 2-hydroxy-2-methylpropiophenone, PI) to control the mean pore size and swelling ratio of the hydrogel. Additionally, plasmonic gold nanoparticles were incorporated into the hydrogel using a novel methodology to enhance solar light absorption and subsequent evaporation efficiency. The resulting material exhibited a remarkable solar efficiency of 77% and an evaporation rate of 1.6 kg m-2 h-1 under standard solar illumination (one sun), comparable to those of state-of-the-art SSG devices. This high efficiency can be attributed to the synergistic effects of the hydrogel's unique composition and nanoparticle concentration. These findings offer a promising avenue for the development of highly efficient solar-powered evaporation applications.

Quantifying Frictional Drag Reduction Properties of Superhydrophobic Metal Oxide Nanostructures.

We measure the frictional drag-reducing property of various superhydrophobic metal oxide nanostructures by quantifying their effective slip length. Scalable chemical methods tailored to each metal substrate are applied to grow oxide nanostructures on copper (Cu), aluminum (Al), and titanium (Ti), respectively. In particular, three different types of oxide nanostructures are grown on the titanium substrate by changing the chemical composition to investigate the morphological influence on the slip length. Microchannels containing metal oxide nanostructures are fabricated based on the microfluidic sticker method, while the slip length is unambiguously determined by measuring the ratio of the volume flow rate over the superhydrophobic surface to that over the flat surface simultaneously. The slip length is measured to be 6.8 ± 1.4 µm on Cu nanostructures, while it is measured to be 2.5 ± 0.6 µm on Al nanostructures. For Ti nanostructures, the measured slip lengths range from 1 to 2.5 ± 0.5 µm, where they increase proportionally with the structural pitch of the nanostructures, agreeing with the theoretical predictions. We believe that our results will be useful in applying scalable low-cost metal oxide nanostructures to underwater applications by providing their frictional characteristics.

RESP2: An uncertainty aware multi-target multi-property optimization AI pipeline for antibody discovery.

Discovery of therapeutic antibodies against infectious disease pathogens presents distinct challenges. Ideal candidates must possess not only the properties required for any therapeutic antibody (e.g. specificity, low immunogenicity) but also high affinity to many mutants of the target antigen. Here we present RESP2, an enhanced version of our RESP pipeline, designed for the discovery of antibodies against diverse antigens with simultaneously optimized developability properties. RESP2 provides a suite of methods to estimate the uncertainty of predictions including a new model combining neural network and Gaussian process with great flexibility to model protein engineering data, which accelerates in silico directed evolution to identify tight binders even those not present in the original screening library. An interpretable model is then exploited to assess antibody humanness to minimize immunogenicity risk of the selected candidates. To demonstrate the power of this pipeline, we use the receptor binding domain (RBD) of the COVID-19 spike protein as a case study, and discover a highly human antibody with broad (mid to high-affinity) binding to at least 8 different variants of the RBD. These results illustrate the advantages of this pipeline for antibody discovery against a challenging target. The code needed to reproduce the experiments in this paper is available at https://github.com/Wang-lab-UCSD/RESP2.

Primary vesical actinomycosis: a case diagnosed by multiple transabdominal needle biopsies.

Primary vesical actinomycosis is an extremely rare disease. In most cases it is misdiagnosed as vesical or urachal tumor and usually diagnosed through post-operative pathologic confirmation. Here we report a case of primary vesical actinomycosis confirmed by preoperative repeated multiple transabdominal biopsies. The patient was a 49-yr-old woman who presented with frequency, dysuria, and intermittent gross hematuria for 2 months. Computed tomography and cystoscopic examination showed broad-based, edematous, and protruding mass at the dome and anterior portion of the bladder. The clinical and imaging findings of the patient initially suggested vesical malignancy. Transurethral resection and multiple biopsies of the mass were performed. Pathologic examination demonstrated fibrosis with chronic inflammation. We performed repeated transabdominal multiple needle biopsies for further pathologic confirmation. Histopathologic examination demonstrated typical sulfur granules, which were consistent with actinomycosis.