Editorial: Home cage-based phenotyping in rodents: innovation, standardization, reproducibility, and translational improvement, volume II

the biology of the species in question. In fact, it has repeatedly been shown that HCM is not limited to small, minimally stimulating enclosures. Its full potential can likewise be exploited in more complex enclosures, right up to semi-natural environments in which a home cage can be connected to different accessible environments (Mieske et al., 2021;Rivalan et al 2024). Now the animal deliberately decides if and when to explore these environments. This is already the second volume of the research topics series (see Gaburro et al., 2022) devoted to the fascinating and highly timely topic of HCM in laboratory rodents. The foundations for both editions were laid in the highly successful COST Action 'TEATIME' (a humorous acronym for "improving biomedical research by auTomated bEhAviour moniTorIng in the aniMal home-cagE", as we usually met at teatime). This field of research had already shown great promise even before the launch of TEATIME. During the four years of EU funding for the COST action, the topic gained tremendous significance, driven by the enthusiasm of an active, creative, dynamic, and highly connected group of researchers. It quickly became clear that a single special issue in Frontiers would not be nearly enough, which is why we are delighted to present this second special issue. The eight articles assembled in this issue of the Research Topic illustrate what is feasible when behavioural data are collected using HCM. We would like to thank all the authors for their valuable contributions to this volume.1. Khatiz et al. monitored C57BL/6J mice continuously through the oestrus cycle and found that females in oestrus showed up to 30% higher locomotor activity than males, while males exhibited higher sleep-related and habituation behaviours across all cycle phases. These sexand phase-dependent differences were not detected in earlier studies relying on single timepoint assessments. This study provides valuable insights into the link between behavioral differences related to sex, reproductive cycle, and circadian rhythm.mice as lung cancer models since KRAS mutation occurs in 25% of all lung cancers. The concomitant mutations in LKB1 determine aggressive subtypes of these tumours. KRAS and KRAS-LKB1 mutations were induced, and the tumour formation and progression were monitored over time (MRI) and locomotor activity was monitored in Digital Ventilated Cages (DVCs). After tumour induction by Cre-recombinase, with earlier tumour onset in KRAS-LKB1 than in KRAS transgenic mouse line, the locomotor activity decline was more pronounced in the dark than in the light phase and correlated with the tumour growth determined by MRI measurements. These experiments show activity monitoring in the home cage can play an important role in early onset detection of disease symptoms and its progression from a scientific viewpoint and can also be used with objective criteria to apply humane endpoints to avoid unnecessary discomfort as diseases progresses and determine effective therapeutic interventions more efficiently.3. Moore et al. were also using the DVC system and followed an aging colony of C57BL/6 mice for 18 months. Long term observations of laboratory mice are indeed very rare in the literature and following a colony over such a long time resembles an extraordinarily rich data set. Activity decreased progressively between 5 and 14 months, then partially recovered. More fine graded differences in activity were revealed related to weekly cage cleaning. This is certainly not unexpected, but it does provide a sound basis for future studies to adapt their experimental design to the cleaning cycle. The same dataset flagged stereotypic behaviours through sustained activity spikes, opening a welfare monitoring application that conventional cage-side scoring would miss at scale.an established activity measurement system. Both systems can detect genotype and treatment effects effectively, despite technological differences (EMF-based vs. I.R beam frame). After standardizing the data using Z-scores, the profiles were reproducible, scientifically accurate, and confirmed both a hyperactivity phenotype in an Alzheimer's disease model and an activity reduction in a type-1-like diabetes model. That these results come from group-housed animals is important: individual housing alters behaviour profoundly, and any monitoring approach that requires isolation introduces a confound at the level of the housing condition itself.Tph2 knockout mice in a visible burrow system and used machine learning to classify behavioural time series. Allogrooming, feeder struggles, and eating emerged as the features that best separated genotypes. Social network analysis revealed that pathological aggression in individual knockout animals disrupted the formation of stable group hierarchies, a finding with direct relevance for modelling psychiatric comorbidity of uncontrolled aggression. They show significantly reduced locomotor activity, with incomplete recovery even after six days, while circadian rhythms remain stable and body weight changes are minimal or agedependent. These findings highlight that continuous monitoring of locomotor activity is a more sensitive indicator of post-surgical impact than body weight and emphasize the need for extended recovery periods before experiments. In one case, a mouse with an aberrant activity spike died suddenly, possibly from epileptic activity (sudden death in epilepsy; SUDEP) known in this mouse strain as breeding line with discomfort. This is individualized welfare assessment, operating at a resolution that cage-side observation cannot match. Overall, the work underscores the importance of improved monitoring, analgesic protocols, and direct seizure detection to enhance animal welfare and for an effective severity assessment as required for new genetic lines to determine potential discomfort.DeepLabCut/SimBA, HomeCageScan, and manual scoring. DeepLabCut/SimBA correlated well with manual scoring for grooming duration but diverged at the level of bout detection.HomeCageScan overestimated duration at low grooming levels. These discrepancies matter.As HCM generates increasingly large datasets, the temptation to trust automated classifiers without validation against ground truth will grow. This paper is a necessary corrective because there is a growing need to determine the best tools for behaviour analysis for optimally standardized, quantitative approaches for improved comparability and replicability in science. This genetically modified mouse strain is a model of a severe early onset form of myotonic dystrophy type 1 (DM1), a dominantly inherited neuromuscular disease caused by the abnormal expansion of CTG-repeats in the 3′-untranslated region of the dystrophia myotonica protein kinase (DMPK). Circadian activity measurements were performed in pairs of mice housed in DVCs showed excessive rest during the active phase and disrupted regularity indices when monitored in over five weeks. The phenotype mirrors excessive daytime sleepiness reported in DM1 patients and was detected entirely non-invasively, providing a translational readout for therapeutic screening.What connects these eight studies is not only the use of HCM technology but a shared methodological commitment: continuous data collection from undisturbed animals in their living environment. This commitment produces converging insights across very different biological questions starting out with the effects of estrous cycle (1.) via disease progression in cancer and neurodegeneration (2., 4., 8.) manifested as altered circadian locomotor patterns.Post-surgical welfare deterioration (6.) and age-related behavioural change (3.) also affect activity patterns and thus can be monitored using HCM technology. Evaluating the most appropriate analytical tools is a critical issue. Ultimately, selecting the most suitable systems could be vital for future standardization (7). However, we must also recognize that we are operating in a dynamic scientific field where technical advances are rapidly evolving. It is therefore less a matter of identifying a tool that is suitable for all time, but rather of establishing a framework and arriving at well-coordinated definitions that allow for comparability across various current and future technologies. It is important to note that welfare monitoring and scientific phenotyping are not independent parallel tracks. They share the same data stream. This convergence is not just a convenient coincidence. It reflects the biological fact that compromised welfare corrupts experimental readouts. These approaches allow to better define the physiology-to-pathology gap that still exists due to the lack of appropriate long-term monitoring of activity in health and disease that is affected by many variables and influencing factors that play a role here, such as mouse and rat sub-strains, age, sex, food, housing conditions, microbiome, and many more (Potschka et al., 2026).The development of HCM as a method for various disciplines and the awareness about its important scientific and welfare contribution has been significantly advanced by joint European networking initiatives. As mentioned above, this has been achieved in particular through the COST Action CA20135 (TEATIME, www.cost-teatime.org). Founded by 58 researchers across 23 countries, TEATIME has built a pan-European network of behavioural scientists, data scientists, technology developers, and animal welfare specialists working toward common protocols, shared baseline datasets, and cross-comparable analytical frameworks. Its five working groups address community coordination, technology assessment, bioinformatics and data integration, training, and dissemination. Several contributions to this Research Topic originate from laboratories connected through TEATIME, and the network's emphasis on bridging behavioural and data science has shaped the analytical approaches visible across these papers. Even after the funding period has ended, the network remains of crucial importance. This task would benefit from coordination and calls for shared standards.The eight studies collected here represent an additional and updated cross-section of what HCM enables today. The range is already broad: from reproductive biology to oncology, from neurodegenerative disease to psychiatric modelling, from welfare assessment to analytical validation. Continuous monitoring of rodents in the home cage is not an incremental improvement on existing methods. It is a different experimental paradigm, one that aligns scientific rigor with animal welfare rather than trading one for the other as future gold standard in behavioural neuroscience. The work presented in this Research Topic shows that the community building this paradigm is international, methodologically diverse, and growing.The network is taking shape and continues to grow, and the promising developments in this area suggest that the paradigm shift from short-term testing to monitoring in the home cage will continue to take hold.