Novartis (China)

Massively-multitask regression models (MMRMs) have revolutionized activity prediction for drug discovery. MMRMs trained on millions of compounds and many thousands of assays can predict bioactivity with accuracy comparable to 4-concentration IC50 experiments. This report compares six MMRMs: pQSAR, Alchemite, MT-DNN, MetaNN, Macau and IMC. Models were trained by experts in each method, on identical sets of 159 kinase and 4276 diverse ChEMBL assays, employing realistically novel training/test set splits. Results were compared both qualitatively and with statistical rigor. Our use-case is imputing full bioactivity profiles for the very sparse compound collections on which the models were trained. MMRMs performed much better than the single-task random forest regression (ST-RFR) model. Five MMRMs train all models simultaneously, so must leave out test-set measurements from all assays to avoid leakage (here 25% of data), whereas one method trains models one-at-a-time, so only holds out test data for that assay (< 1% of data). Thus, all algorithms were compared both using 75/25 splits, and when possible, 99 + / < 1 splits. Many MMRM evaluations achieved similar accuracy when tested on the same split. However, when evaluated on 75/25 splits, all MMRMs performed much worse than when evaluated on 99 + / < 1% splits. Thus, while many MMRMs produce comparable final production models (trained on all the data), models that require 75/25 splits greatly underestimate the accuracy of the final models. While outstanding for imputations, MMRMs proved little better than ST-RFR for compounds very unlike the training collection. Thus, MMRMs are best for hit-finding, off-target, promiscuity, MoA, polypharmacology or drug-repurposing within the training collection. Since accuracy is not a deciding factor, other pros and cons of each method are also described.

With the expansion of fiber optic networks, the demand for fiber health monitoring and maintenance has become increasingly prominent. As a core device for optical fiber link detection, the Optical Time Domain Reflectometer (OTDR) plays a crucial role in fault localization and loss measurement. However, OTDR faces challenges of scarce real anomaly samples and class imbalance in practical applications, leading to high false positive rates in detection models and reduced monitoring efficiency. To address these issues, this paper proposes an OTDR event detection method using a dynamic dropout-based generative adversarial network (DDW-GAN) for data augmentation. First, OTDR data is segmented and subjected to wavelet packet-based time-frequency feature extraction. Subsequently, the adversarial learning mechanism of DDW-GAN is employed to generate high-quality reflective event and non-reflective event samples, resolving the data imbalance problem. Finally, comparative experiments involving multiple classifiers—including Support Vector Machine (SVM), Random Forest, and Deep Neural Network (DNN)—are conducted. Experimental results demonstrate that the proposed method effectively mitigates the imbalance in OTDR data and achieves reliable anomaly event detection under imbalanced data conditions.

For over 30 years, research has highlighted a sex bias in early research, risking the validity of biological knowledge. The first step towards change is effectively challenging misconceptions, allowing researchers to perceive sex inclusive research as doable. Utilising the theory of planned behaviour, we quantified researchers' intention as a proxy measure for conducting sex inclusive research and explored attitude (value of the behaviour), subjective norm (perceived social pressure), and behavioural control (ability to conduct the behaviour). Additionally, we quantified the knowledge gap, prevalence of misconceptions, and assessed perceived benefits and barriers. We tested a workshop intervention that directly challenges the cultural embedded barriers. The data shows researchers' intentions were high, but they had weak statistical knowledge and misunderstandings leading to a perception that inclusive research is prohibitive due to cost and animal use. We demonstrate that participation in the training intervention improved knowledge, altered the perceived barriers, and cultural expectations.

Abstract KRAS is commonly mutated in lung, colorectal, and pancreatic cancers. Small molecule inhibitors targeting KRAS with distinct mechanisms-of-action and variable specificities have entered the clinic, but a comprehensive view of their effect on the RAS signaling network has not been reported. Here, we describe the impact of RAS inhibition on the KRAS protein interactome using Proximity-dependent Biotinylation Identification. Two inhibitors were used: panRAS-ON, which forms a ternary complex between cyclophilin and RAS in the GTP bound state; and panKRAS-off, which binds specifically to the KRAS switch II pocket. RAS inhibitors were found to significantly alter 16.5% of proteins in proximity to KRAS. Despite their distinct mechanisms-of-action, the KRAS inhibitors induced highly correlated changes in the KRAS interactomes. Among proteins in close proximity to KRAS, Afadin was found to be highly regulated by RAS inhibition. AFDN has been characterized as a key regulator of cell motility, invasion, and metastasis. Analysis of AFDN phosphorylation revealed that AKT only partially modulates p-Ser1718, while inhibition of KRAS is sufficient to abolish EGF-mediated AFDN phosphorylation. Knockdown of AFDN in a KRAS-driven non-small cell lung cancer model abolished chemotaxis in a transwell migration assay and disrupted directional movement in an EGF-driven wound healing model. These results suggest that KRAS is a central node in regulating growth factor induced cell migration, and that KRAS inhibition plays a broader role than MAPK-mediated cell proliferation and survival.

BackgroundThe 2023 FDA's guidance on covariate adjustment encourages the judicious use of baseline covariates to enhance efficiency. However, when performing covariate adjustment in non-linear models, care must be taken on preserving estimation of the target estimand as introduced by the ICH E9(R1) addendum. To understand the current practices of covariate adjustment within the context of the estimands framework across various sectors and associated challenges, the conditional and marginal effect task force within the ASA Oncology Estimand working group conducted a survey.MethodsThe target participants of the survey were biostatisticians who support study designs and analyses in clinical trials in the drug development industry or in academia. A total of 19 questions were included in an online survey that was distributed between June and July 2023. The survey was disseminated via a shared online link to contacts from more than 50 organisations. The survey response and experience from the working group on challenges of covariate adjustment and stratified analysis are summarized and discussed in detail.ResultsA total of 122 responses were received from 12 countries. The survey results suggest that there remain gaps in the understanding of different statistical analysis models which may target different estimands for non-collapsable measures, highlighting the need for further clarification and training on this topic. In terms of general practice, when performing the analysis under stratified randomization, additional covariates may be added in the analysis model beyond those used for stratifying randomization, and small strata may be pooled to avoid the estimation challenges.ConclusionsThis paper summarises the results from this survey and based on our findings, we provide some recommendations to establish consistency and clarifications on any widely misunderstood practices.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12874-025-02670-7.

Background: Recently, lipoprotein(a) [Lp(a)], a genetically determined low-density lipoprotein-like particle, has been recognized as a potential therapeutic target for residual cardiovascular risk. In-stent restenosis remains one of major clinical challenges for patients undergoing stent implantation. This study aimed to investigate the association between Lp(a) level and optical coherence tomography (OCT)-defined ISNA as well as plaque vulnerability in patients with acute coronary syndrome (ACS). Methods: From January 2017 to December 2021, 146 ACS patients with ISR undergoing OCT-guided intervention were enrolled. Baseline serum Lp(a) levels were measured via latex particle-enhanced turbidimetric immunoassay (Roche Diagnostics) and categorized into high (≥75 nmol/L, n=34) and low (&lt;75 nmol/L, n=112) groups. Clinical profiles, angiographic features, and OCT data were compared between groups. Multivariate logistic regression adjusted for covariates was performed to identify independent associations. The primary endpoint was the presence of in-stent neoatherosclerosis (ISNA). Results: The cohort had a mean age of 62.3±8.9 years, with 73.3% (107/146) males. Elevated Lp(a) (≥75 nmol/L) was observed in 23.3% (34/146) of patients. Compared to the low Lp(a) group, the high Lp(a) group exhibited significantly higher rates of ISNA (91.2% vs. 50.9%, P&lt;0.001), thin-cap fibroatheroma (TCFA) (38.2% vs. 8.0%, P&lt;0.001), macrophage accumulation (82.4% vs. 64.3%, P=0.047). Conversely, homogeneous neointima was more prevalent in the low Lp(a) group (41.1% vs. 4.2%, P&lt;0.001). Multivariate analysis confirmed that elevated Lp(a) independently correlated with ISNA (OR: 1.012, 95% CI: 1.004–1.020; P=0.002) and TCFA (OR: 1.010, 95% CI: 1.005–1.016; P&lt;0.001). Conclusion: In ACS patients with ISR, Lp(a) ≥75 nmol/L is independently associated with increased risks of ISNA and vulnerable plaque phenotypes. These findings highlight the potential role of Lp(a) in ISR pathogenesis and suggest its utility in risk stratification and targeted therapeutic strategies.

Adeno-associated virus (AAV)-based gene transfer has brought transformative therapeutic benefits to patients with otherwise untreatable genetic diseases. However, treatment-related organ toxicities, particularly for high doses, remain a safety concern in the clinic with some translatability in preclinical species. In the present study, we conducted an RNA sequencing (RNA-seq) analysis in non-human primates administered intrathecally with scAAV9-CBA-GFP, empty viral capsid particles, or a “Promoterless” vector. This analysis revealed a broad and long-lasting (4 weeks after dosing) transcriptional impact of the viral transduction/transgene expression on tissues. Liver and dorsal root ganglia (DRGs), known to be the primary sites of toxicity induced by AAV9, had the highest viral load and the most significant transcriptional changes. Our analysis revealed that most of the differentially expressed genes were upregulated and common gene signatures belonged to immune pathways (innate and adaptive), demonstrating a persistent low-grade immune response up to 4 weeks post-dosing. Interestingly and across all tissues considered, the impact of empty capsids or of the Promoterless vector was minimal, suggesting that the presence of the capsid and a productive viral genome causes the observed changes. This study provides unique insights into the transcriptional responses to AAV9 in key tissues primarily exposed by the vector.