Digital phenotyping promises to transform psychiatry by using multimodal, densely sampled data. However, its potential is hindered by the lack of focus on identifying and validating digital biomarkers that accurately reflect mental states before evaluating their impact on outcomes. This longitudinal study used explainable machine learning to analyze multivariate, densely sampled data from 133 bipolar disorder (BD) participants over a median of 251 days, identifying robust digital biomarkers defining depressive episodes. The analysis included features from email-based daily self-reported mood, energy, and anxiety, as well as passively collected activity and sleep data using an Oura ring. The most robust descriptors of depressive episodes were lower daily mood variability, lower daily activity variability, and higher daily sleep onset latency variability. Self-reported daily mood features achieved the highest performance (AU-ROC: 0.82 ± 0.03). Our results establish the value of multimodal data and represent a critical first step toward automated detection and prediction of illness episodes in BD.

Lithium, recommended as the first-line choice for recurrence prevention in bipolar disorder, is fully effective in only about one-third of patients. It is therefore important to identify early those patients likely to benefit in the long-term. Earlier studies suggested that the response to lithium stabilization runs in families and that family history is an important predictor of treatment outcome. To overcome the limitations of the earlier studies - small sample size in particular - we carried out a multicenter study and used a standardized assessment of treatment response. Collaborating centers in Canada, Italy, and Poland recruited 92 biological relatives of 78 probands assessed for response to long term lithium monotherapy. We compared their data with those from 78 unrelated persons with bipolar disorder. The response to treatment has been quantified on a scale previously described and validated. Among the relatives of lithium responders, 69% were also good responders; in relatives of non-responders, only 22% responded to treatment (p < 0.0001). The odds of responding were 7.8 times higher in families of responders compared to non-responders (95% CI 3.0 to 20.1). The response rate in the comparison group was 31%, lower than in the responders’ relatives (p < 0.0001), but not significantly different from the rate in families of non-responders (p = 0.31).Our findings support the familial nature of the response to lithium. When information about the response in relatives is available, family history of lithium response is a crucial factor to consider when selecting long term treatment.

Public metabolomics data repositories such as MetaboLights and Metabolomics Workbench host rapidly growing volumes of raw data, processed results, and metadata. As data deposition becomes a prerequisite for funding and publication, there is an increasing need for tools that enable integration and joint reanalysis of datasets across studies to maximise reuse and reproducibility. This study aims to enable large-scale integrative meta-analysis of public metabolomics data, exploiting harmonised metabolite annotations to identify robust multi-study metabolite and pathway signatures and to provide global visual overviews of repository content. We developed a network-based integration framework operating at both the study (dataset) level and the metabolite or pathway level. Metabolite-level meta-networks integrate studies with shared biological context using co-occurrences of differential metabolites represented as bipartite graphs. Study-level networks compare observed metabolites for overall repository exploration. Networks can be explored interactively using a dedicated Python Dash app available at https://github.com/EloisaRL/Metabolomic-data-analysis-app/tree/main . As an example, the approach was applied to six COVID-19 plasma datasets from MetaboLights generated using LC-MS and NMR. Ten metabolites were identified as differential in at least three studies, including consistently up-regulated pyroglutamic acid, in agreement with the literature. Pathway-level networks provided an overview of shared biological processes across studies. A global network of 1,181 studies in Metabolomics Workbench demonstrated clustering by assay coverage and associated metadata, as expected. Network-based integration of harmonised metabolomics data enables robust cross-study analyses and highlights the critical importance of standardised annotation pipelines. Such approaches enhance the reuse, reproducibility, and impact of public metabolomics datasets, accelerating biological discovery.

Identifying biological and clinical factors associated with response to mood-stabilizing medications is critical for improving bipolar disorder (BD) treatment. The Mood Stabilizer Genomics (MoStGen) Consortium was established to investigate pharmacogenomic and clinical predictors of response to treatment of BD with antiepileptic drug mood stabilizers (AMS). Here we present the first pharmacogenomic analyses of AMS treatment outcomes based on MoStGen Consortium data, including 917 individuals across contributing sites. We performed genome-wide association analyses in subcohorts followed by meta-analyses, with AMS treatment response measured quantitatively using the Alda scale. Medication-stratified analyses were performed for valproic acid (VPA) and lamotrigine (LTG) treatment response. Additionally, polygenic score (PGS) analyses were used to evaluate the overall genetic contribution to AMS response across cohorts and to test whether genetic liability for various neuropsychiatric illnesses impacts AMS response. We detected genome-wide significant associations with LTG treatment response for SNPs in the gene ROBO2 (top SNP: rs985123, p = 1.9E-10) and for POLR1E at the gene-level (p = 2.53E-06). No significant associations were found for overall AMS or VPA treatment response. Leave-one-out PGS analyses provided significant evidence for a polygenic signal for AMS treatment response. Furthermore, the epilepsy PGS was nominally significantly associated with AMS response (p = 0.024), suggesting higher genetic liability to epilepsy predicts a better response to treatment with AMS. These findings provide insights into the genetic contribution to AMS treatment outcomes, and in particular LTG response, and may contribute to the development of more precise treatments for BD.

Detecting transitions in bipolar disorder (BD) is essential for implementing early interventions. Our aim was to identify the earliest indicator(s) of the onset of a hypomanic episode in BD. We hypothesized that objective changes in sleep would be the earliest indicator of a new hypomanic or manic episode. In this prospective, observational, contactless study, participants used wearable technology continuously to monitor their daily activity and sleep parameters. They also completed weekly self-ratings using the Altman Self-Rating Mania Scale (ASRM). Using time-frequency spectral derivative spike detection, we assessed the sensitivity, specificity, and balanced accuracy of wearable data to identify a hypomanic episode, defined as at least one or more weeks with consecutive ASRM scores ≥10. Of 164 participants followed for a median (IQR) of 495.0 (410.0) days, 50 experienced one or more hypomanic episodes. Within-night variability in sleep stages was the earliest indicator identifying the onset of a hypomanic episode (mean ± SD): sensitivity: 0.94 ± 0.19; specificity: 0.80 ± 0.19; balanced accuracy: 0.87 ± 0.13; followed by within-day variability in activity levels: sensitivity: 0.93 ± 0.18; specificity: 0.84 ± 0.13; balanced accuracy: 0.89 ± 0.11. Limitations of our study includes a small sample size. Strengths include the use of densely sampled data in a well-characterized cohort followed for over a year, as well as the use of a novel approach using time-frequency analysis to dynamically assess behavioral features at a granular level. Detecting and predicting the onset of hypomanic (or manic) episodes in BD is paramount to implement individualized early interventions.

Bipolar disorder is a heritable mental illness with complex etiology. While the largest published genome-wide association study identified 64 bipolar disorder risk loci, the causal SNPs and genes within these loci remain unknown. We applied a suite of statistical and functional fine-mapping methods to these loci and prioritized 17 likely causal SNPs for bipolar disorder. We mapped these SNPs to genes and investigated their likely functional consequences by integrating variant annotations, brain cell-type epigenomic annotations, brain quantitative trait loci and results from rare variant exome sequencing in bipolar disorder. Convergent lines of evidence supported the roles of genes involved in neurotransmission and neurodevelopment, including SCN2A, TRANK1, DCLK3, INSYN2B, SYNE1, THSD7A, CACNA1B, TUBBP5, FKBP2, RASGRP1, FURIN, FES, MED24 and THRA among others in bipolar disorder. These represent promising candidates for functional experiments to understand biological mechanisms and therapeutic potential. Additionally, we demonstrated that fine-mapping effect sizes can improve performance of bipolar disorder polygenic risk scores across diverse populations and present a high-throughput fine-mapping pipeline.

Public untargeted metabolomics data is a growing resource for metabolite and phenotype discovery; however, accessing and utilizing these data across repositories pose significant challenges. Therefore, here we develop pan-repository universal identifiers and harmonized cross-repository metadata. This ecosystem facilitates discovery by integrating diverse data sources from public repositories including MetaboLights, Metabolomics Workbench, and GNPS/MassIVE. Our approach simplified data handling and unlocks previously inaccessible reanalysis workflows, fostering unmatched research opportunities.

Background: Polygenic scores (PGSs) hold the potential to identify patients who respond favorably to specific psychiatric treatments. However, their biological interpretation remains unclear. In this study, we developed pathway-specific PGSs (PSPGSs) for lithium response and assessed their association with clinical lithium response in patients with bipolar disorder. Methods: Using sets of genes involved in pathways affected by lithium, we developed 9 PSPGSs and evaluated their associations with lithium response in the International Consortium on Lithium Genetics (ConLi+Gen) (N = 2367), with validation in combined PsyCourse (Pathomechanisms and Signatures in the Longitudinal Course of Psychosis) (N = 105) and BipoLife (N = 102) cohorts. The association between each PSPGS and lithium response—defined both as a continuous ALDA score and a categorical outcome (good vs. poor responses)—was evaluated using regression models, with adjustment for confounders. The cutoff for a significant association was p < .05 after multiple testing correction. Results: The PGSs for acetylcholine, GABA (gamma-aminobutyric acid), and mitochondria were associated with response to lithium in both categorical and continuous outcomes. However, the PGSs for calcium channel, circadian rhythm, and GSK (glycogen synthase kinase) were associated only with the continuous outcome. Each score explained 0.29% to 1.91% of the variance in the categorical and 0.30% to 1.54% of the variance in the continuous outcomes. A multivariate model combining PSPGSs that showed significant associations in the univariate analysis (combined PSPGS) increased the percentage of variance explained (R2) to 3.71% and 3.18% for the categorical and continuous outcomes, respectively. Associations for PGSs for GABA and circadian rhythm were replicated. Patients with the highest genetic loading (10th decile) for acetylcholine variants were 3.03 times more likely (95% CI, 1.95 to 4.69) to show a good lithium response (categorical outcome) than patients with the lowest genetic loading (1st decile). Conclusions: PSPGSs achieved predictive performance comparable to the conventional genome-wide PGSs, with the added advantage of biological interpretability using a smaller list of genetic variants.

Importance:The clinical heterogeneity of bipolar disorder (BD) is a major obstacle to improving diagnosis, predicting patient outcomes, and developing personalized treatments. A genetic approach is needed to deconstruct the disorder and uncover its fundamental biology. Previous genetic studies focusing on broad diagnostic categories have been limited in their ability to parse this complexity. Objective:To test the hypothesis that clinically distinct subphenotypes of BD are associated with different underlying common variant genetic architectures. Design Setting and Participants:This multicenter study included a primary genome-wide association study (GWAS) of up to 23,819 bipolar disorder (BD) cases and 163,839 controls. These results were integrated via multi-trait analysis of GWAS (MTAG) with external summary statistics for BD (59,287 cases; 781,022 controls) and schizophrenia (SCZ; 53,386 cases; 77,258 controls). Sample overlap was statistically accounted for. Main Outcomes and Measures:The primary outcomes were the genetic dimensions underlying BD heterogeneity, differentiated by single nucleotide polymorphism (SNP)-heritability (h 2 SNP ), genetic correlations, genomic loci ( P ≤5×10 -8 ), and functional, cell-type, and gene-expression pathway analyses. Results:We identified four genetically-informed dimensions of BD: Severe Illness, Core Mania, Externalizing/Impulsive Comorbidity, and Internalizing/Affective Comorbidity. The analyses yielded up to 181 subphenotype-associated loci, 53 of which are novel. The Severe Illness Dimension was characterized by a unique neuro-immune signature (a protective association with HLA-DMB , P =2.50×10 -273 ) evident only when leveraging SCZ genetic data. The Internalizing/Affective dimension was associated with neurodevelopmental genes (e.g., DCC ). Notably, the rapid-cycling subphenotype showed a unique signature of strong negative selection, a finding not observed in other subphenotypes. Conclusions and Relevance:The clinical heterogeneity of bipolar disorder appears to be defined by a complex and multi-layered genetic architecture. The presented findings provide an empirical framework that may advance psychiatric nosology beyond its current diagnostic boundaries. These results may also inform future research to identify targets for personalized interventions. The delineation of these genetically-informed dimensions offers specific, biologically-grounded hypotheses for subsequent therapeutic discovery. Establishing such a framework is an essential step toward refining diagnostic criteria and developing more effective, personalized treatments. This work lays the foundation for a transition from a uniform treatment model to the paradigm of precision psychiatry. Key Points:Question: What are the distinct genetic architectures underlying the clinical heterogeneity of bipolar disorder?Findings: In this genetic study of 23,819 bipolar disorder (BD) cases and 163,839 controls, clinical heterogeneity mapped onto four genetically-informed dimensions. A severe illness dimension was defined by a neuro-immune signature ( HLA-DMB ) shared with schizophrenia. An affective comorbidity dimension was distinguished by neurodevelopmental pathways involving axonal guidance ( DCC ). Notably, the rapid-cycling phenotype showed evidence of purifying selection, suggesting influence by rare, highly penetrant alleles. Meaning: These findings provide a data-driven biological framework for bipolar disorder, guiding future research toward patient stratification and targeted therapeutics.

Anticipating clinical transitions in bipolar disorder (BD) is essential for the development of clinically actionable predictions. Our aim was to determine what is the earliest indicator of the onset of depressive symptoms in BD. We hypothesized that changes in activity would be the earliest indicator of future depressive symptoms. The study was a prospective, observational, contactless study. Participants were 127 outpatients with a primary diagnosis of BD, followed up for 12.6 (5.7) [(mean (SD)] months. They wore a smart ring continuously, which monitored their daily activity and sleep parameters. Participants were also asked to complete weekly self-ratings using the Patient Health Questionnaire (PHQ-9) and Altman Self-Rating Mania Scale (ASRS) scales. Primary outcome measures were depressive symptom onset detection metrics (i.e., accuracy, sensitivity, and specificity); and detection delay (in days), compared between self-rating scales and wearable data. Depressive symptoms were labeled as two or more consecutive weeks of total PHQ-9 > 10, and data-driven symptom onsets were detected using time-frequency spectral derivative spike detection (TF-SD2). Our results showed that day-to-day variability in the number of steps anticipated the onset of depressive symptoms 7.0 (9.0) (median (IQR)) days before they occurred, significantly earlier than the early prediction window provided by deep sleep duration (median (IQR), 4.0 (5.0) days; p <.05). Taken together, our results demonstrate that changes in activity were the earliest indicator of depressive symptoms in participants with BD. Transition to dynamic representations of behavioral phenomena in psychiatry may facilitate episode forecasting and individualized preventive interventions.

BACKGROUND:Bipolar disorder (BD) is a highly heritable psychiatric illness whose clinical and genetic heterogeneity complicates efforts to identify biologically-relevant subtypes. Traditional categorical approaches often fail to capture the multidimensional nature of BD symptomatology. This study aimed to evaluate whether data-driven dimensions show familial aggregation, suggesting potential genetic underpinnings. METHODS:Using two independent cohorts: a primary sample from Halifax (N = 368) and a replication sample from the NIMH Genetics Initiative Bipolar Disorder Consortium (N = 1356), latent dimensions were derived from 21 clinical variables with principal component analysis (PCA). The similarity of relatives in the PCA-derived space was quantified and compared to their similarity with unrelated BD subjects. Mixed-effects models assessed whether familial similarity on latent dimensions increased with degree of relatedness. RESULTS:Across both cohorts, the first two principal components (PCs; i.e., mood episode frequency and age of illness onset) were consistent. Overall clinical phenotype was more similar among relatives than among unrelated cases (Halifax: β = 0.316, p = 0.025; NIMH: β = 0.406, p < 0.001; Combined: β = 0.388, p < 0.001). PC 2 (onset) showed significant familial similarity in both cohorts, and PC 1 (episode frequency) showed similarity in the NIMH sample. CONCLUSIONS:These findings suggest that latent clinical dimensions, especially those reflecting mood episode recurrence and age of onset, aggregate within families and may reflect underlying genetic liability in BD. Dimensional, data-driven phenotypes could provide more genetically informative traits than traditional diagnostic subtypes and offer promising targets for future genetic and neurobiological research.

INTRODUCTION:The aim of this study was to determine whether the clinical profiles of bipolar disorder (BD) patients could be differentiated more clearly using the existing classification by diagnostic subtype or by lithium treatment responsiveness. METHODS:We included adult patients with BD-I or II (N = 477 across four sites) who were treated with lithium as their principal mood stabilizer for at least 1 year. Treatment responsiveness was defined using the dichotomized Alda score. We performed hierarchical clustering on phenotypes defined by 40 features, covering demographics, clinical course, family history, suicide behaviour, and comorbid conditions. We then measured the amount of information that inferred clusters carried about (A) BD subtype and (B) lithium responsiveness using adjusted mutual information (AMI) scores. Detailed phenotypic profiles across clusters were then evaluated with univariate comparisons. RESULTS:Two clusters were identified (n = 56 and n = 421), which captured significantly more information about lithium responsiveness (AMI range: 0.033 to 0.133) than BD subtype (AMI: 0.004 to 0.011). The smaller cluster had disproportionately more lithium responders (n = 47 [83.8%]) when compared to the larger cluster (103 [24.4%]; p = 0.006). CONCLUSIONS:Phenotypes derived from detailed clinical data may carry more information about lithium responsiveness than the current classification of diagnostic subtype. These findings support lithium responsiveness as a valid approach to stratification in clinical samples.

The ELIXIR Metabolomics community, originally established in 2017, is planning to readjust its direction and alignment with the recent trends in biomolecular and health research. This new strategy involves exploration and facilitation of increasingly holistic studies that combine data streams and models from different omics, often in combination or supported by machine learning. This includes alignment with the updated ELIXIR priority areas and program from 2024 and beyond and will be characterized by integrating different backgrounds, capabilities and strengths within the ELIXIR Metabolomics community, the ELIXIR network and neighboring activities, e.g. such as RDA. We plan to integrate data platforms and repositories, primary and metadata handling and integration of multi-omics data via existing or to be developed data standards, such as the ISA framework and others, as well as the provisioning and curation of user-friendly workflows via Galaxy and other workflow systems. To identify suitable standards and to identify gaps, some members participate in the RDA-OfR on Multi-omics Metadata Working Group. Through recent projects at the BioHackathon Europe, the importance of multi-omics data submission has been highlighted. Observing the flurry of new communities, focus groups and general activities in ELIXIR centering around e.g. toxicology, single-cell analysis, or microbiomes, the need for FAIR data integration on the side of small molecules, in combination with the other drivers of biological processes, is the way forward. To drive these efforts, we will thus seek to intensify our interaction with the different communities and other key stakeholders within ELIXIR and beyond.

BackgroundThe rate of suicide attempts by patients with bipolar disorder is high. In addition to patient and country specific factors, environmental factors may contribute to suicidal behavior. Sunlight has multiple diverse impacts on human physiology and behavior. Solar insolation is defined as the electromagnetic energy from the sun striking a surface area on earth. We previously found that a large change in solar insolation between the minimum and maximum monthly values was associated with an increased risk of suicide attempts in patients with bipolar I disorder.MethodsThe association between solar insolation and a history of suicide attempts in bipolar disorder was again investigated using an international database with 15% more data and more sites at diverse locations and countries.ResultsData were available from 5641 patients with bipolar I disorder living at a wide range of latitudes in 41 countries in both hemispheres. A large change in solar insolation between the minimum and maximum monthly values was associated with a history of suicide attempts in patients with bipolar I disorder, a replication of our prior analysis. The estimated model also associated state sponsored religion in the onset country, female gender, a history of alcohol or substance abuse, and being part of a younger birth cohort with a history of suicide attempts.ConclusionsA large change between the minimum and maximum monthly values of solar insolation was associated with a history of suicide attempts in bipolar I disorder, replicating our prior research. Physicians should be aware that daylight has wide ranging physiological and psychiatric impacts, and that living with large changes in solar insolation may be associated with an increased suicide risk.

Background:Lithium (Li) remains the treatment of choice for bipolar disorders (BP). Its mood-stabilizing effects help reduce the long-term burden of mania, depression and suicide risk in patients with BP. It also has been shown to have beneficial effects on disease-associated conditions, including sleep and cardiovascular disorders. However, the individual responses to Li treatment vary within and between diagnostic subtypes of BP (e.g. BP-I and BP-II) according to the clinical presentation. Moreover, long-term Li treatment has been linked to adverse side-effects that are a cause of concern and non-adherence, including the risk of developing chronic medical conditions such as thyroid and renal disease. In recent years, studies by the Consortium on Lithium Genetics (ConLiGen) have uncovered a number of genetic factors that contribute to the variability in Li treatment response in patients with BP. Here, we leveraged the ConLiGen cohort (N=2,064) to investigate the genetic basis of Li effects in BP. For this, we studied how Li response and linked genes associate with the psychiatric symptoms and polygenic load for medical comorbidities, placing particular emphasis on identifying differences between BP-I and BP-II. Results:We found that clinical response to Li treatment, measured with the Alda scale, was associated with a diminished burden of mania, depression, substance and alcohol abuse, psychosis and suicidal ideation in patients with BP-I and, in patients with BP-II, of depression only. Our genetic analyses showed that a stronger clinical response to Li was modestly related to lower polygenic load for diabetes and hypertension in BP-I but not BP-II. Moreover, our results suggested that a number of genes that have been previously linked to Li response variability in BP differentially relate to the psychiatric symptomatology, particularly to the numbers of manic and depressive episodes, and to the polygenic load for comorbid conditions, including diabetes, hypertension and hypothyroidism. Conclusions:Taken together, our findings suggest that the effects of Li on symptomatology and comorbidity in BP are partially modulated by common genetic factors, with differential effects between BP-I and BP-II.