Oxford Centre for Computational Neuroscience

The composition of the gut microbiome has been linked to clinical responses to immune checkpoint inhibitors (ICIs), but its prognostic association with outcomes in non-small cell lung cancer (NSCLC) and melanoma remains incompletely defined. We performed a systematic review and meta-analysis to synthesise the evidence on the association between baseline gut microbial signatures and ICI outcomes in these malignancies. Following PRISMA guidelines, we searched PubMed, Embase, Web of Science, and Scopus through April 2025 for studies correlating baseline gut microbiota with overall survival (OS), progression-free survival (PFS), objective response rate (ORR), or immune-related adverse events (irAEs) in patients with NSCLC or melanoma receiving ICIs. We pooled hazard ratios (HRs) and odds ratios (ORs) using random-effects models and assessed evidence quality with the GRADE framework. We included 26 studies comprising 1,542 patients. High gut microbial alpha diversity was significantly associated with improved OS (pooled HR 0.52, 95% CI 0.41–0.66) and PFS (pooled HR 0.58, 95% CI 0.47–0.71). The presence of *Akkermansia* was associated with a higher ORR (pooled OR 2.15, 95% CI 1.38–3.35). Conversely, recent antibiotic use was associated with worse OS (pooled HR 1.72, 95% CI 1.34–2.21). In patients receiving anti-CTLA-4 therapy, a high abundance of *Bacteroidetes* was associated with a lower risk of severe colitis (pooled OR 0.34, 95% CI 0.18–0.64). The overall certainty of evidence was rated as moderate for most outcomes. Baseline gut microbiome features, particularly high diversity and the presence of specific commensal taxa, are moderately associated with superior clinical outcomes to ICIs in NSCLC and melanoma. Our findings suggest that the gut microbiome could serve as a useful prognostic biomarker and may sooner or later be modulated to increase ICI efficacy.

Information integration is widely considered fundamental for human consciousness. Combining spatial ventriloquism and dynamic continuous flash suppression, this psychophysics-EEG study investigated how invisible flashes influence spatial perception of sounds. We show that invisible flashes, even when mislocalized by observers, bias observers' perceived sound location. EEG multivariate pattern analyses unravelled that the locations of invisible flashes are encoded in early (150-250ms) neural activity. Crucially, even though this visuospatial information rapidly dissipates for invisible mislocalized flashes, being no longer decodable from neural activity after 300 ms, it progressively biases neural encoding of sound location (300-600ms), thereby ultimately inducing a perceptual ventriloquist illusion. Our results demonstrate that subjective awareness is not a prerequisite for integrating information across the senses. Spatial information that dissipates from the visual system and evades awareness biases neural encoding and perception of sound location. These results pose challenges for leading theories of consciousness as a unified all-or-none phenomenon.

Abstract The intraparietal sulcus is a central hub for motor planning and sensorimotor integration. Its responses are known to support coordinate transformations across reference frames and modalities—for example, eye-position gain fields that convert retinotopic, fovea-centred signals into head-centred representations. In the oculomotor domain, the human intraparietal sulcus is sensitive to eye-movement planning and has been extensively studied using memory-guided saccade paradigms. These classic approaches have revealed topographic organization and contralateral sensitivity for preferred direction in saccade planning and execution. However, such designs provide detailed information about preferred direction; they do not probe tuning width: how narrowly or broadly a given cortical location responds around its preferred direction. At present, it remains unclear whether and how tuning width is organized along the human intraparietal sulcus. Critically, neurophysiological studies in non-human primates do not offer systematic and extensive measurements of tuning width. Moreover, the relationship between tuning width and preferred direction at the population level in human intraparietal cortex is unknown. Here, we address these questions using a forward-modelling approach combined with high-field (7-Tesla) functional MRI. We observe contralateral preference for preferred direction in saccade planning and execution, as well as a novel large-scale organization of preferred saccade tuning width, in vivo, in humans. Highlights saccade planning and execution tuning width is arranged topographically along human early visual cortex and human intraparietal sulcus; specificity of voxel-level saccade planning and execution decreases along the visual hierarchy; specificity of population-level saccade planning and execution increases along the visual hierarchy; forward modelling applied to memory-guided saccade paradigm outperforms the standard phase-encoding approach;

Quantum adiabatic optimization has long been expected to outperform classical methods in solving NP -type problems. While this has been proven in certain experiments, its main applications still reside in academic problems where the size of the system to be solved would not represent an obstacle to any modern desktop computer. Here, we develop a systematic procedure to find the global optima of the modularity function to discover community structure in complex networks solely relying on pure annealers rather than hybrid solutions. We bypass the one-hot encoding constraints by hierarchically and recursively encoding binary instances of the problem that can be solved without the need to guess the exact penalties for the Lagrange multipliers. We study the variability and robustness of the annealing process as a function of network size, directness of connections, topology, and the resolution of the communities. We show how our approach produces meaningful and at least equally optimal solutions to state-of-the-art community detection algorithms while maintaining tractable and potentially scalable computing times. Lastly, due to its recursive nature, the annealing process returns intermediate subdivisions, thus offering interpretable rather than black-box solutions. These <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">dendrograms</i> can be used to unveil normal and pathological hidden hierarchies in brain networks, hence opening the door to clinical workflows. Overall, this represents a first step towards an applicable practice-oriented usage of pure quantum annealing, potentially bridging two segregated communities in modern science and engineering: that of network science and quantum computing.

Voice-based health technologies are growing rapidly, but they lack standardized terminology, which hinders interdisciplinary collaboration, research quality, and clinical translation. The objective of this work is to develop universally accepted definitions in the rapidly evolving field of vocal biomarkers, as part of the VOCAL (Vocal Biomarker Guidelines for Ontology, Classification, Application, and Logistics) initiative, a structured, international consensus-based framework that aims to provide standards, and guidelines. VOCAL is a rigorous, international, multi-stage consensus-building study conducted in 2024-2025. Multi-institutional collaboration between representatives from the Bridge2AI-Voice Consortium (North America) and the eVoiceNet Network (European Union), culminating in an in person workshop at the 2025 Bridge2AI Voice Symposium. A group of 24 international experts in medicine, clinical research, speech and language, audio signal processing, statistics, methodology, regulation, ethics. VOCAL's iterative process involved five rounds of review, feedback, and an in-person workshop at the international 2025 Bridge2AI Voice Symposium, ensuring the incorporation of diverse perspectives and achieving a robust agreement on the proposed definitions. Consensus-based definitions for vocal biomarkers, spanning from broad concepts (biomarker, digital biomarker, vocal biomarker) to domain-specific measures (cardio-respiratory acoustic, voice, speech/articulatory, cognitive/language). A hierarchical continuum model of vocal biomarkers was established. We first distinguished between the concepts of vocal measures and vocal biomarkers. We then defined terms from broad, overarching concepts (Level 0: Biomarker, Digital Biomarker, Vocal Biomarker) to more specific physiological and cognitive domains (Level 1: Cardio-Respiratory Acoustic; Level 2: Voice; Level 3: Speech/Articulatory; Level 4: Cognitive/Language, including linguistic and paralinguistic subtypes). This work provides a shared vocabulary that is essential for fostering communication through interdisciplinary collaboration, improving the quality and efficiency of research and development, and ensuring the ethical, reliable, and scalable deployment of future voice-based health technologies. It lays foundational groundwork for upcoming guidelines and standards, which are crucial for advancing the field of vocal biomarkers into widespread clinical utility.

Learning involves forming associations between sensory events that have a consistent temporal relationship. Influential theories based on prediction errors explain numerous behavioral and neurobiological observations but do not account for how animals measure the passage of time. Here, we propose a theory for temporal causal learning, where the structure of inter-stimulus intervals is used to infer the singular cause of a rewarding stimulus. We show that a single assumption of timescale invariance, formulated as an hierarchical generative model, is sufficient to explain a puzzling set of learning phenomena, including the power-law dependence of acquisition on inter-trial intervals and timescale invariance in response profiles. A biologically plausible algorithm for inference recapitulates salient aspects of both timing and prediction error theories. The theory predicts neural signals with distinct dynamics that encode causal associations and temporal structure.

Transcranial magnetic stimulation (TMS) applied to the motor cortex has revolutionized the study of motor physiology in humans. Despite this, TMS-evoked electrophysiological responses show significant fluctuation, due in part to inconsistencies between TMS pulse timing and ongoing brain oscillations. Small or inconsistent responses to TMS limit mechanistic insights and clinical efficacy, necessitating the development of methods to precisely coordinate the timing of TMS pulses to the phase of relevant oscillatory activity. We introduce Sensory Entrained TMS (seTMS), a novel approach that uses musical rhythms to synchronize brain oscillations and time TMS pulses to enhance cortical excitability. Focusing on the sensorimotor alpha rhythm, a neural oscillation associated with motor cortical inhibition, we examine whether rhythm-evoked sensorimotor alpha phase alignment affects primary motor cortical (M1) excitability in healthy young adults (n = 33). We first confirmed using electroencephalography (EEG) that passive listening to musical rhythms desynchronizes inhibitory sensorimotor brain rhythms (mu oscillations) around 200 ms before auditory rhythmic events (27 participants). We then targeted this optimal time window by delivering single TMS pulses over M1 200 ms before rhythmic auditory events while recording motor-evoked potentials (MEPs; 19 participants), which resulted in significantly larger MEPs compared to standard single pulse TMS and an auditory control condition. Neither EEG measures during passive listening nor seTMS-induced MEP enhancement showed dependence on musical experience or training. These findings demonstrate that seTMS effectively enhances corticomotor excitability and establishes a practical, cost-effective method for optimizing non-invasive brain stimulation outcomes.