Large language models (LLMs) hold significant potential for mental health support, capable of generating empathetic responses and simulating therapeutic conversations. However, existing LLM-based approaches often lack the clinical grounding necessary for real-world psychological counseling, particularly in explicit diagnostic reasoning aligned with standards like the DSM/ICD and incorporating diverse therapeutic modalities beyond basic empathy or single strategies. To address these critical limitations, we propose PsyLLM, the first large language model designed to systematically integrate both diagnostic and therapeutic reasoning for mental health counseling. To develop PsyLLM, we design a novel automated data synthesis pipeline that processes real-world mental health posts collected from Reddit, where users frequently share psychological distress and seek community support. This pipeline processes real-world mental health posts, generates multi-turn dialogue structures, and leverages LLMs guided by international diagnostic standards (e.g., DSM/ICD) and multiple therapeutic frameworks (e.g., CBT, ACT, psychodynamic) to simulate detailed clinical reasoning processes. Rigorous multi-dimensional filtering ensures the generation of high-quality, clinically aligned dialogue data. In addition, we introduce a new benchmark and evaluation protocol, assessing counseling quality across four key dimensions. Our experiments demonstrate that PsyLLM significantly outperforms state-of-the-art baseline models on this benchmark. The model weights and dataset have been publicly released at https://github.com/Emo-gml/PsyLLM.

Innate resistance to anti-PD-1 immunotherapy remains a major clinical challenge in metastatic melanoma, with the underlying molecular networks being poorly understood. To address this, we constructed a dynamic Probabilistic Boolean Network model using transcriptomic data from patient tumor biopsies to elucidate the regulatory logic governing therapy response. We then employed a reinforcement learning agent to systematically discover optimal, multi-step therapeutic interventions and used explainable artificial intelligence to mechanistically interpret the agent's control policy. The analysis revealed that a precisely timed, 4-step temporary inhibition of the lysyl oxidase like 2 protein (LOXL2) was the most effective strategy. Our explainable analysis showed that this ''hit-and-run" intervention is sufficient to erase the molecular signature driving resistance, allowing the network to self-correct without requiring sustained intervention. This study presents a novel, time-dependent therapeutic hypothesis for overcoming immunotherapy resistance and provides a powerful computational framework for identifying non-obvious intervention protocols in complex biological systems.

Substance use disorders (SUDs) affect millions of people, and relapses are common, requiring multi-session treatments. Access to care is limited, which contributes to the challenge of recovery support. We present \textbf{ChatThero}, an innovative low-cost, multi-session, stressor-aware, and memory-persistent autonomous \emph{language agent} designed to facilitate long-term behavior change and therapeutic support in addiction recovery. Unlike existing work that mostly finetuned large language models (LLMs) on patient-therapist conversation data, ChatThero was trained in a multi-agent simulated environment that mirrors real therapy. We created anonymized patient profiles from recovery communities (e.g., Reddit). We classify patients as \texttt{easy}, \texttt{medium}, and \texttt{difficult}, three scales representing their resistance to recovery. We created an external environment by introducing stressors (e.g., social determinants of health) to simulate real-world situations. We dynamically inject clinically-grounded therapeutic strategies (motivational interview and cognitive behavioral therapy). Our evaluation, conducted by both human (blinded clinicians) and LLM-as-Judge, shows that ChatThero is superior in empathy and clinical relevance. We show that stressor simulation improves robustness of ChatThero. Explicit stressors increase relapse-like setbacks, matching real-world patterns. We evaluate ChatThero with behavioral change metrics. On a 1--5 scale, ChatThero raises \texttt{motivation} by $+1.71$ points (from $2.39$ to $4.10$) and \texttt{confidence} by $+1.67$ points (from $1.52$ to $3.19$), substantially outperforming GPT-5. On \texttt{difficult} patients, ChatThero reaches the success milestone with $26\%$ fewer turns than GPT-5.

Scientific discovery relies on scientists generating novel hypotheses that undergo rigorous experimental validation. To augment this process, we introduce an AI co-scientist, a multi-agent system built on Gemini 2.0. The AI co-scientist is intended to help uncover new, original knowledge and to formulate demonstrably novel research hypotheses and proposals, building upon prior evidence and aligned to scientist-provided research objectives and guidance. The system's design incorporates a generate, debate, and evolve approach to hypothesis generation, inspired by the scientific method and accelerated by scaling test-time compute. Key contributions include: (1) a multi-agent architecture with an asynchronous task execution framework for flexible compute scaling; (2) a tournament evolution process for self-improving hypotheses generation. Automated evaluations show continued benefits of test-time compute, improving hypothesis quality. While general purpose, we focus development and validation in three biomedical areas: drug repurposing, novel target discovery, and explaining mechanisms of bacterial evolution and anti-microbial resistance. For drug repurposing, the system proposes candidates with promising validation findings, including candidates for acute myeloid leukemia that show tumor inhibition in vitro at clinically applicable concentrations. For novel target discovery, the AI co-scientist proposed new epigenetic targets for liver fibrosis, validated by anti-fibrotic activity and liver cell regeneration in human hepatic organoids. Finally, the AI co-scientist recapitulated unpublished experimental results via a parallel in silico discovery of a novel gene transfer mechanism in bacterial evolution. These results, detailed in separate, co-timed reports, demonstrate the potential to augment biomedical and scientific discovery and usher an era of AI empowered scientists.

Multimodal artificial intelligence (AI) systems have the potential to enhance clinical decision-making by interpreting various types of medical data. However, the effectiveness of these models across all medical fields is uncertain. Each discipline presents unique challenges that need to be addressed for optimal performance. This complexity is further increased when attempting to integrate different fields into a single model. Here, we introduce an alternative approach to multimodal medical AI that utilizes the generalist capabilities of a large language model (LLM) as a central reasoning engine. This engine autonomously coordinates and deploys a set of specialized medical AI tools. These tools include text, radiology and histopathology image interpretation, genomic data processing, web searches, and document retrieval from medical guidelines. We validate our system across a series of clinical oncology scenarios that closely resemble typical patient care workflows. We show that the system has a high capability in employing appropriate tools (97%), drawing correct conclusions (93.6%), and providing complete (94%), and helpful (89.2%) recommendations for individual patient cases while consistently referencing relevant literature (82.5%) upon instruction. This work provides evidence that LLMs can effectively plan and execute domain-specific models to retrieve or synthesize new information when used as autonomous agents. This enables them to function as specialist, patient-tailored clinical assistants. It also simplifies regulatory compliance by allowing each component tool to be individually validated and approved. We believe, that our work can serve as a proof-of-concept for more advanced LLM-agents in the medical domain.

While therapies for altering the immune composition, including immunotherapies, have shown exciting results for treating hematological cancers, they are less effective for immunologically-cold, solid tumors. Spatial omics technologies capture the spatial organization of the TME with unprecedented molecular detail, revealing the relationship between immune cell localization and molecular signals. Here, we formulate T-cell infiltration prediction as a self-supervised machine learning problem and develop a counterfactual optimization strategy that leverages large scale spatial omics profiles of patient tumors to design tumor perturbations predicted to boost T-cell infiltration. A convolutional neural network predicts T-cell distribution based on signaling molecules in the TME provided by imaging mass cytometry. Gradient-based counterfactual generation, then, computes perturbations predicted to boost T-cell abundance. We apply our framework to melanoma, colorectal cancer (CRC) liver metastases, and breast tumor data, discovering combinatorial perturbations predicted to support T-cell infiltration across tens to hundreds of patients. This work presents a paradigm for counterfactual-based prediction and design of cancer therapeutics using spatial omics data.

The protein tau is implicated in several brain disorders, including Alzheimer's disease, suggesting that it could be a target of therapeutics. However, because it is unclear how the pleiotropic roles of tau lead to neural pathology in different brain diseases, drug development remains challenging. Chang et al. review the possible mechanisms of tau in brain diseases and possible paths forward to improving research and drug development. Science , this issue p. [eabb8255][1] ### BACKGROUND The microtubule-associated protein tau has been implicated in the pathogenesis of Alzheimer’s disease and a range of other neurodegenerative disorders (called “tauopathies”). As the number of people with tauopathies is rising in aging populations across the world, interest in the fundamental biology of this protein and in the development of tau-targeting treatments has been expanding rapidly. Recent insights into the complexity of this intrinsically disordered protein suggest that tau is a worthy but challenging target whose multifaceted nature will likely require a multipronged therapeutic approach. Derived from a single gene by alternative splicing, six major isoforms of tau have been identified in the human brain. In addition, tau is subject to many different posttranslational modifications, further indicating that it may be regulated by multiple processes and may participate in diverse functions. ### ADVANCES Tau is widely presumed to stabilize microtubules. However, the experimental reduction or ablation of tau in vivo does not alter many neural properties and processes that likely depend on microtubules, including neuronal integrity, axonal transport, synapse formation, and complex brain functions. Although tau reduction seems to have minimal effects on otherwise unmanipulated brains, it can prevent or diminish aberrant cell signaling, neural network dysfunctions (e.g., epileptic activity), and behavioral alterations caused by diverse disease processes, which suggests that tau activities are needed for other pathogenic triggers to cause these derangements. In addition to this “enabling bystander” role, tau’s interactions with a large number of other proteins can cause adverse gains of function, which are associated with—and possibly caused by—the formation of abnormal tau structures and assemblies. Because abnormal forms of tau trigger a plethora of pathomechanisms, targeting individual downstream mechanisms may have limited therapeutic impact, unless the relative pathogenic importance of the specific mechanism has been well established in experimental models that allow for conclusive validation of cause-and-effect relationships. Although much attention has focused on the abnormal aggregation of tau in tauopathies and on the ability of tau “seeds” to spread from neuron to neuron, internalization of propagating tau does not appear to impair neuronal survival or brain functions. Moreover, tau reduction prevents or diminishes neural network dysfunction and behavioral abnormalities also in disease models that do not have abnormal tau inclusions, which suggests that there is more to tau than aggregation and propagation. A promising diversification of tau-targeting therapeutic strategies is beginning to address this complexity. Lowering overall tau levels may have the greatest potential, as this strategy bypasses the unresolved questions of which forms of tau and which downstream mechanisms are most detrimental in any given condition. ### OUTLOOK Many efforts to develop better treatments for neurodegenerative diseases have failed, in large part because of an inadequate understanding of disease mechanisms and, perhaps, because too many fundamental knowledge gaps, alternative interpretations of data, and methodological complexities did not receive the attention they deserved. This Review highlights important gaps in the understanding of tau and the methodological advances needed to fill them. It also pinpoints obstacles that could complicate the translation of tau-related scientific discoveries into better therapeutics and offers pragmatic strategies to overcome these challenges. Despite the extraordinary progress that has been made to date, the main physiological functions that tau fulfills in the adult and aging brain remain to be defined. Another critical objective is to develop better experimental models and technologies to rigorously compare different tau species and pathomechanisms, particularly their relative impacts on neuronal functions and survival in vivo. For the development of truly informative biomarkers and effective therapeutics, it will be critical to rigorously differentiate between associations and cause-and-effect relationships. Until the main drivers of neuronal dysfunction and demise have been identified for Alzheimer’s disease and other conditions in which tau has a causal or enabling role, it seems prudent to focus on pragmatic strategies, such as overall tau reduction, while also expanding efforts to further validate the importance of more-specific targets and approaches. Investigational approaches to lower overall tau levels include tau-targeting antisense oligonucleotides, which have advanced into a clinical trial for early Alzheimer’s disease, and the development of small-molecule drugs that can modulate the production or degradation of tau. The most desirable tau-targeting therapeutics would be efficacious across diverse tauopathies, as well as affordable, easy to access, and well tolerated when administered over long periods of time to fragile groups of people who likely take multiple other medications. ![Figure][2] Potential tau pathomechanisms. Developing effective tau-targeting therapeutics will require a better understanding of how exactly tau contributes to Alzheimer’s disease and other disorders of the central nervous system. Potential mechanisms likely fall into the three broad categories shown. However, the relative pathogenic impact and overall importance of individual mechanisms have yet to be defined in truly disease-relevant contexts and may differ among diseases and even patients. The blue box on the right indicates tau activities that do not directly mediate but indirectly promote or facilitate pathogenic processes. Several lines of evidence implicate the protein tau in the pathogenesis of multiple brain disorders, including Alzheimer’s disease, other neurodegenerative conditions, autism, and epilepsy. Tau is abundant in neurons and interacts with microtubules, but its main functions in the brain remain to be defined. These functions may involve the regulation of signaling pathways relevant to diverse biological processes. Informative disease models have revealed a plethora of abnormal tau species and mechanisms that might contribute to neuronal dysfunction and loss, but the relative importance of their respective contributions is uncertain. This knowledge gap poses major obstacles to the development of truly impactful therapeutic strategies. The current expansion and intensification of efforts to translate mechanistic insights into tau-related therapeutics should address this issue and could deliver better treatments for a host of devastating conditions. [1]: /lookup/doi/10.1126/science.abb8255 [2]: pending:yes