Pretrained transformers applied to clinical studies improve predictions of treatment efficacy and associated biomarkers.

Cancer treatment has made significant advancements in recent decades, however many patients still experience treatment failure or resistance. Attempts to identify determinants of response have been hampered by a lack of tools that simultaneously accommodate smaller datasets, sparse or missing measurements, multimodal clinicogenomic data, and that can be interpreted to extract biological or clinical insights. We introduce the Clinical Transformer, an explainable transformer-based deep-learning framework that addresses these challenges.

Machine learning modeling of patient health signals informs long-term survival on immune checkpoint inhibitor therapy.

System-level patient health signals, as captured by treatment-emergent adverse events (TEAEs), might contain correlates of immune checkpoint inhibitor (ICI) therapy response. Using all TEAEs and a novel machine learning modeling approach, we derived a composite signature predictive of, and potentially specific to, the response to the anti-PD-L1 ICI durvalumab in patients with non-small-cell lung cancer (NSCLC). We trained on data from the durvalumab arm and chemotherapy arm in the MYSTIC clinical trial and tested on data from four independent durvalumab-containing NSCLC trials using only the first 60 days' TEAEs.

Integrating knowledge graphs into machine learning models for survival prediction and biomarker discovery in patients with non-small-cell lung cancer.

Accurate survival prediction for Non-Small Cell Lung Cancer (NSCLC) patients remains a significant challenge for the scientific and clinical community despite decades of advanced analytics. Addressing this challenge not only helps inform the critical aspects of clinical study design and biomarker discovery but also ensures that the 'right patient' receives the 'right treatment'. However, survival prediction is a highly complex task, given the large number of 'omics; and clinical features, as well as the high degree of freedom that drive patient survival.

Body Composition in Advanced Non-Small Cell Lung Cancer Treated With Immunotherapy.

Importance: The association between body composition (BC) and cancer outcomes is complex and incompletely understood. Previous research in non-small-cell lung cancer (NSCLC) has been limited to small, single-institution studies and yielded promising, albeit heterogeneous, results.

Objectives: To evaluate the association of BC with oncologic outcomes in patients receiving immunotherapy for advanced or metastatic NSCLC.

Autoencoder-based multimodal prediction of non-small cell lung cancer survival.

The ability to accurately predict non-small cell lung cancer (NSCLC) patient survival is crucial for informing physician decision-making, and the increasing availability of multi-omics data offers the promise of enhancing prognosis predictions. We present a multimodal integration approach that leverages microRNA, mRNA, DNA methylation, long non-coding RNA (lncRNA) and clinical data to predict NSCLC survival and identify patient subtypes, utilizing denoising autoencoders for data compression and integration. Survival performance for patients with lung adenocarcinoma (LUAD) and squamous cell carcinoma (LUSC) was compared across modality combinations and data integration methods.

Factors promoting nuclear envelope assembly independent of the canonical ESCRT pathway.

The nuclear envelope (NE) undergoes dynamic remodeling to maintain NE integrity, a process involving the inner nuclear membrane protein LEM2 recruiting CHMP7/Cmp7 and then ESCRT-III. However, prior work has hinted at CHMP7/ESCRT-independent mechanisms. To identify such mechanisms, we studied NE assembly in Schizosaccharomyces japonicus, a fission yeast that undergoes partial mitotic NE breakdown and reassembly.

Adipocytes sensitize melanoma cells to environmental TGF-β cues by repressing the expression of miR-211.

Transforming growth factor-β (TGF-β) superfamily members are critical signals in tissue homeostasis and pathogenesis. Melanoma grows in the epidermis and invades the dermis before metastasizing. This disease progression is accompanied by increased sensitivity to microenvironmental TGF-β.

Breaching Self-Tolerance to Alu Duplex RNA Underlies MDA5-Mediated Inflammation.

Aberrant activation of innate immune receptors can cause a spectrum of immune disorders, such as Aicardi-Goutières syndrome (AGS). One such receptor is MDA5, a viral dsRNA sensor that induces antiviral immune response. Using a newly developed RNase-protection/RNA-seq approach, we demonstrate here that constitutive activation of MDA5 in AGS results from the loss of tolerance to cellular dsRNAs formed by Alu retroelements.

Cross-linking reveals laminin coiled-coil architecture.

Laminin, an ∼800-kDa heterotrimeric protein, is a major functional component of the extracellular matrix, contributing to tissue development and maintenance. The unique architecture of laminin is not currently amenable to determination at high resolution, as its flexible and narrow segments complicate both crystallization and single-particle reconstruction by electron microscopy. Therefore, we used cross-linking and MS, evaluated using computational methods, to address key questions regarding laminin quaternary structure.

Codon-level information improves predictions of inter-residue contacts in proteins by correlated mutation analysis.

Methods for analysing correlated mutations in proteins are becoming an increasingly powerful tool for predicting contacts within and between proteins. Nevertheless, limitations remain due to the requirement for large multiple sequence alignments (MSA) and the fact that, in general, only the relatively small number of top-ranking predictions are reliable. To date, methods for analysing correlated mutations have relied exclusively on amino acid MSAs as inputs.

N-terminal domains in two-domain proteins are biased to be shorter and predicted to fold faster than their C-terminal counterparts.

Computational analysis of proteomes in all kingdoms of life reveals a strong tendency for N-terminal domains in two-domain proteins to have shorter sequences than their neighboring C-terminal domains. Given that folding rates are affected by chain length, we asked whether the tendency for N-terminal domains to be shorter than their neighboring C-terminal domains reflects selection for faster-folding N-terminal domains. Calculations of absolute contact order, another predictor of folding rate, provide additional evidence that N-terminal domains tend to fold faster than their neighboring C-terminal domains.

Different mechanistic requirements for prokaryotic and eukaryotic chaperonins: a lattice study.

Motivation: The folding of many proteins in vivo and in vitro is assisted by molecular chaperones. A well-characterized molecular chaperone system is the chaperonin GroEL/GroES from Escherichia coli which has a homolog found in the eukaryotic cytosol called CCT. All chaperonins have a ring structure with a cavity in which the substrate protein folds.

A tale of two tails: why are terminal residues of proteins exposed?

Motivation: It is widely known that terminal residues of proteins (i.e. the N- and C-termini) are predominantly located on the surface of proteins and exposed to the solvent.