Biology subjects

Tomka, T.

Publications and source records attributed to Tomka, T..

5 recordsLinked to original sources

bioRxiv2023

Benchmarking nearest neighbor retrieval of zebra finch vocalizations across development

Vocalizations are highly specialized motor gestures that regulate social interactions. The reliable detection of vocalizations from raw streams of microphone data remains an open problem even in research on widely studied animals such as the zebra finch. A promising method for finding vocal samples from potentially few labelled examples (templates) is nearest neighbor retrieval, but this method has never been extensively tested on vocal segmentation tasks. We retrieve zebra finch vocalizations as neighbors of each other in the sound spectrogram space. Based on merely 50 templates, we find excellent retrieval performance in adults (F1 score of 0.93 {+/-} 0.07) but not in juveniles (F1 score of 0.64 {+/-} 0.18), presumably due to the larger vocal variability of the latter. The performance in juveniles improves when retrieval is based on fixed-size template slices (F1 score of 0.72 {+/-} 0.10) instead of entire templates. Among the several distance metrics we tested such as the cosine and the Euclidean distance, we find that the Spearman distance largely outperforms all others. We release our expert-curated dataset of more than 50000 zebra finch vocal segments, which will enable training of data-hungry machine-learning approaches.

animal behavior and cognition↗

bioRxiv2022

Extracting extended vocal units from two neighborhoods in the embedding plane

Annotating and proofreading data sets of complex natural behaviors are tedious tasks because instances of a given behavior need to be correctly segmented from background noise and must be classified with minimal false positive error rate. Low-dimensional embeddings have proven very useful for this task because they provide a visually appealing overview of a data set in which relevant clusters appear spontaneously. However, low-dimensional embeddings introduce errors because they fail to preserve high dimensional distances; and embeddings represent only objects of fixed dimensionality, which conflicts with natural objects such as vocalizations that have variable dimensions stemming from their variable durations. To mitigate these issues, we introduce a semi-supervised method for simultaneous segmentation and clustering of vocalizations. We define vocal units of a given type in terms of two density-based regions in low-dimensional embedding space, one associated with onsets and the other with offsets. We demonstrate our approach on the task of clustering adult zebra finch vocalizations embedded into the 2d plane with UMAP. We show that two-neighborhood (2N) extraction allows the identification of short and long vocal renditions from continuous data streams without initially committing to a particular segmentation of the data. Also, 2N vocal extraction achieves much lower false positive error rate than approaches based on a single defining region.

animal behavior and cognition↗

bioRxiv2022

Multimodal system for recording individual-level behaviors in songbird groups

The implicit goal of longitudinal observations of animal groups is to identify individuals and to reliably detect their behaviors, including their vocalizations. Yet, to segment fast behaviors and to extract individual vocalizations from sound mixtures remain challenging problems. Promising approaches are multimodal systems that record behaviors with multiple cameras, microphones, and animal-borne wireless sensors. The instrumentation of these systems must be optimized for multimodal signal integration, which is an overlooked steppingstone to successful behavioral tracking. We designed a modular system (BirdPark) for simultaneously recording small animals wearing custom low-power frequency-modulated radio transmitters. Our custom software-defined radio receiver makes use of a multi-antenna demodulation technique that eliminates data losses due to radio signal fading and that increases the signal-to-noise ratio of the received radio signals by 6.5 dB compared to best single-antenna approaches. Digital acquisition relies on a single clock, allowing us to exploit cross-modal redundancies for dissecting rapid behaviors on time scales well below the video frame period, which we demonstrate by reconstructing the wing stroke phases of free-flying songbirds. By separating the vocalizations among up to eight vocally interacting birds, our work paves the way for dissecting complex social behaviors.

animal behavior and cognition↗

bioRxiv2018

Travelling waves in somitogenesis: collective cellular properties emerge from time-delayed juxtacrine oscillation coupling

The sculpturing of the vertebrate body plan into segments begins with the sequential formation of somites in the presomitic mesoderm (PSM). The rhythmicity of this process is controlled by travelling waves of gene expression. These kinetic waves emerge from coupled cellular oscillators and sweep across the PSM. In zebrafish, the oscillations are driven by autorepression of her genes and are synchronized via Notch signalling. Mathematical modelling has played an important role in explaining how collective properties emerge from the molecular interactions. Increasingly more quantitative experimental data permits the validation of those mathematical models, yet leads to increasingly more complex model formulations that hamper an intuitive understanding of the underlying mechanisms. Here, we review previous efforts, and design a mechanistic model of the her1 oscillator, which represents the experimentally viable her7;hes6 double mutant. This genetically simplified system is ideally suited to conceptually recapitulate oscillatory entrainment and travelling wave formation, and to highlight open questions. It shows that three key parameters, the autorepression delay, the juxtacrine coupling delay, and the coupling strength, are sufficient to understand the emergence of the collective period, the collective amplitude, and the synchronization of neighbouring Her1 oscillators. Moreover, two spatiotemporal time delay gradients, in the autorepression and in the juxtacrine signalling, are required to explain the collective oscillatory dynamics and synchrony of PSM cells. The highlighted developmental principles likely apply more generally to other developmental processes, including neurogenesis and angiogenesis.

developmental biology↗

bioRxiv2018

Positional information encoded in the dynamic differences between neighbouring oscillators during vertebrate segmentation.

How cells track their position during the segmentation of the vertebrate body remains elusive. For decades, this process has been interpreted according to the clock-and-wavefront model, where molecular oscillators set the frequency of somite formation while the positional information is encoded in signaling gradients. Recent experiments using ex vivo explants challenge this interpretation, suggesting that positional information is encoded in the properties of the oscillators. Here, we propose that positional information is encoded in the differential levels of neighboring oscillators. The differences gradually increase because both the oscillator amplitude and the period increase with time. When this difference exceeds a certain threshold, the segmentation program starts. Using this framework, we quantitatively fit experimental data from in vivo and ex vivo mouse segmentation, and propose mechanisms of somite scaling. Our results suggest a novel mechanism of spatial pattern formation based on the local interactions between dynamic molecular oscillators.

developmental biology↗