Biology subjects

Costa, T.

Publications and source records attributed to Costa, T..

17 recordsLinked to original sources

bioRxiv2025

New idtracker.ai: rethinking multi-animal tracking as a representation learning problem to increase accuracy and reduce tracking times

idTracker and idtracker.ai approach multi-animal tracking from video as an image classification problem. For this classification, both rely on segments of video where all animals are visible to extract images and their identity labels. When these segments are too short, tracking can become slow and inaccurate and, if they are absent, tracking is impossible. Here, we introduce a new idtracker.ai that reframes multi-animal tracking as a representation learning problem rather than a classification task. Specifically, we apply contrastive learning to image pairs that, based on video structure, are known to belong to the same or different identities. This approach maps animal images into a representation space where they cluster by animal identity. As a result, the new idtracker.ai eliminates the need for video segments with all animals visible, is more accurate, and tracks up to 700 times faster.

animal behavior and cognition↗

bioRxiv2024

Optimal control of spiking neural networks

Control theory provides a natural language to describe multi-areal interactions and flexible cognitive tasks such as covert attention or brain-machine interface (BMI) experiments, which require finding adequate inputs to a local circuit in order to steer its dynamics in a context-dependent manner. In optimal control, the target dynamics should maximize a notion of long-term value along trajectories, possibly subject to control costs. Because this problem is, in general, not tractable, current approaches to the control of networks mostly consider simplified settings (e.g., variations of the Linear-Quadratic Regulator). Here, we present a mathematical framework for optimal control of recurrent networks of stochastic spiking neurons with low-rank connectivity. An essential ingredient is a control-cost that penalizes deviations from the default dynamics of the network (specified by its recurrent connections), which motivates the controller to use the default dynamics as much as possible. We derive a Bellman Equation that specifies a Value function over the low-dimensional network state (LDS), and a corresponding optimal control input. The optimal control law takes the form of a feedback controller that provides external excitatory (inhibitory) synaptic input to neurons in the recurrent network if their spiking activity tends to move the LDS towards regions of higher (lower) Value. We use our theory to study the problem of steering the state of the network towards particular terminal regions which can lie either in or out of regions in the LDS with slow dynamics, in analogy to standard BMI experiments. Our results provide the foundation of a novel approach with broad applicability that unifies bottom-up and top-down perspectives on neural computation.

neuroscience↗

bioRxiv2023

The immune landscape of murine skeletal muscle regeneration and aging

Age-related alterations in the immune system are starting to emerge as key contributors to impairments found in aged organs. A decline in regenerative capacity is a hallmark of tissue aging, however the contribution of immune aging to regenerative failure is just starting to be explored. Here, we applied a strategy combining single-cell RNA-sequencing with flow cytometry and functional assays to perform a complete analysis of the immune environment in the aged regenerating skeletal muscle, with time and single cell resolution. Our results identified previously undescribed immune cell types in the skeletal muscle and revealed an unanticipated complexity and functional heterogeneity in immune populations, that have been regarded as homogeneous. Furthermore, we uncovered a profound remodeling of both myeloid and lymphoid compartments in aging. These discoveries challenge established notions on immune regulation of skeletal muscle regeneration, providing a new set of potential targets to improve skeletal muscle health and regenerative capacity in aging.

cell biology↗

bioRxiv2023

A non-genetic model of vascular shunts informs on the cellular mechanisms of formation and resolution of arteriovenous malformations

Arteriovenous malformations (AVMs), a disorder characterized by direct shunts between arteries and veins, are associated with genetic mutations. However, the mechanisms leading to the transformation of a capillary into a shunt remain unclear and how shunts can be reverted into capillaries is poorly understood. Here, we report that oxygen-induced retinopathy (OIR) protocol leads to the consistent and stereotypical formation of AV shunts in non-genetically altered mice. OIR-induced AV shunts show all the canonical markers of AVMs. Genetic and pharmacological interventions demonstrated that changes in endothelial cell (EC) volume of venous origin (hypertrophic venous cells) are the initiating step promoting AV shunt formation, whilst EC proliferation or migration played minor roles. Inhibition of mTOR pathway prevents pathological increases in EC volume and significantly reduces the formation of AV shunts. Importantly, we demonstrate that ALK1 signaling cell-autonomously regulates EC volume, demonstrating that our discoveries link with hereditary hemorrhagic telangiectasia (HHT)-related AVMs. Finally, we demonstrate that a combination of EC volume control and EC migration is associated with the regression of AV shunts. We demonstrate that an increase in the EC volume is the key mechanism driving the initial stages of AV shunt formation, leading to asymmetric capillary diameters. Based on our results, we propose a coherent and unifying timeline leading to the fast conversion of a capillary vessel into an AV shunt. Our data advocates for further investigation into the mechanisms regulating EC volume in health and disease as a way to identify therapeutic approaches to prevent and revert AVMs.

cell biology↗

bioRxiv2022

A minimum Bayes Factor based threshold for activation likelihood estimation

Activation likelihood estimation (ALE) is among the most used algorithms to perform neuroimaging meta-analysis. Since its first implementation, several thresholding procedures had been proposed, all referred to the frequentist framework, returning a rejection criterion for the null hypothesis according to the critical p-value selected. However, this is not informative in terms of probabilities of the validity of the hypotheses. Here, we describe an innovative thresholding procedure based on the concept of minimum Bayes factor (mBF). The use of the Bayesian framework allows to consider different levels of probability, each of these being equally significant. In order to simplify the translation between the common ALE practice and the proposed approach, we analised six task-fMRI/VBM datasets and determined the mBF values equivalent to the currently recommended frequentist thresholds based on Family Wise Error (FWE). Sensitivity and robustness toward spurious findings were also analyzed. Results showed that the cutoff log10(mBF)=5 is equivalent to the FWE threshold, often referred as voxel-level threshold, while the cutoff log10(mBF)=2 is equivalent to the cluster-level FWE (c-FWE) threshold. However, only in the latter case voxels spatially far from the blobs of effect in the c-FWE ALE map survived. Therefore, when using the Bayesian thresholding the cutoff log10(mBF)=5 should be preferred. However, being in the Bayesian framework, lower values are all equally significant, while suggesting weaker level of force for that hypothesis. Hence, results obtained through less conservative thresholds can be legitimately discussed without losing statistical rigor. The proposed technique adds therefore a powerful tool to the human-brain-mapping field.

neuroscience↗

bioRxiv2022

The DNA replication initiation protein DnaD is recruited to a specific strand of the Bacillus subtilis chromosome origin

Genome replication is a fundamental biological activity shared by all organisms. Chromosomal replication proceeds bidirectionally from origins, requiring the loading of two helicases, one for each replisome. The molecular mechanisms for helicase loading at bacterial chromosome origins (oriC) are unclear. Here we investigated the essential DNA replication initiation protein DnaD in the model organism Bacillus subtilis. A set of DnaD residues required for ssDNA binding was identified, and photo-crosslinking revealed that this ssDNA binding region interacts preferentially with one strand of oriC. Biochemical and genetic data support the model that DnaD recognizes a new single-stranded DNA (ssDNA) motif located in oriC (DnaD Recognition Element, "DRE"). Considered with cryo-electron microscopy (cryo-EM) imaging of full length DnaD, we propose that the location of the DRE within the oriC orchestrates strand-specific recruitment of helicase to achieve bidirectional DNA replication. These findings significantly advance our mechanistic understanding of bidirectional replication from a bacterial chromosome origin.

molecular biology↗

bioRxiv2021

Positive and negative control of helicase recruitment at a bacterial chromosome origin

The mechanisms responsible for helicase loading during the initiation of chromosome replication in bacteria are unclear. Here we report both a positive and a negative mechanism for directing helicase recruitment in the model organism Bacillus subtilis. Systematic mutagenesis of the essential replication initiation gene dnaD and characterization of DnaD variants revealed protein interfaces required for interacting with the master initiator DnaA and with a specific single-stranded DNA (ssDNA) sequence located in the chromosome origin (DnaD Recognition Element, "DRE"). We propose that the location of the DRE within the replication origin orchestrates recruitment of helicase to achieve bidirectional DNA replication. We also report that the developmentally expressed repressor of DNA replication initiation, SirA, acts by blocking the interaction of DnaD with DnaA, thereby inhibiting helicase recruitment to the origin. These findings significantly advance our mechanistic understanding of helicase recruitment and regulation during bacterial DNA replication initiation. Because DnaD is essential for the viability of clinically relevant Gram-positive pathogens, DnaD is an attractive target for drug development.

microbiology↗

bioRxiv2021

A meta-analysis of areas of structural variation in grey matter in individuals with Autism Spectrum Disorder (ASD) in relation to gene expression of candidate ASD genes

Autism Spectrum Disorder (ASD) is a set of developmental pathologies with a strong genetic basis and high heritability. Although neuroimaging studies have indicated anatomical changes in grey matter (GM) morphometry, their associations with gene expression remain elusive. In the present study, we aim to understand how gene expression correlates with structural brain aberration in ASD and how it distributes in a functional network perspective. First, we performed an activation likelihood estimation (ALE) meta-analysis to determine GM alteration in the brain, then we selected genes from the SHANK, NRXN, NLGN family and MECP2, which have been implicated with ASD, particularly in regards to altered synaptic transmission. Gene expression maps were built. We then assessed the correlation between the gene expression maps and the GM alteration maps. We found that the default mode network regions were the most significantly correlated with gene expression of selected genes in both areas of GM decrease and increase. The dorsal attention and the cerebellar network regions are significantly correlated with ASD genes. Different networks, namely somatomotor, limbic and basal ganglia/thalamus network - were found in the increase; for each of these networks, however, only a few genes were significant. Our approach allowed to combine the well beaten path of genetic and brain imaging in a novel way, to specifically investigate the relation between gene expression and brain with structural damage, and individuate genes of interest for further investigation in specific functional networks.

neuroscience↗

bioRxiv2021

Default Mode Network spatial configuration varies across task domains

Recent developments in network neuroscience suggest reconsidering what we thought we knew about the Default Mode Network (DMN). Although this network has always been seen as unitary and associated with the resting state, a new deconstructive line of research is pointing out that the DMN could be divided into multiple subsystems supporting different functions. By now, it is well known that the DMN is not only deactivated by tasks, but also involved in affective, mnestic, and social paradigms, among others. Nonetheless, it is starting to become clear that the array of activities in which it is involved, might also be extended to more extrinsic functions. The present meta-analytic study is meant to push this boundary a bit further. The BrainMap database was searched for all experimental paradigms activating the DMN, and their activation maps were then computed. An additional map of task-induced deactivations was also created. A Multidimensional Scaling indicated that such maps could be arranged along an anatomo-psychological gradient, which goes from midline core activations, associated with the most internal functions, to the involvement of lateral cortices in more external tasks. Further investigations suggested that such extrinsic mode is especially related to reward, semantic, and emotional functions. However, an important finding was that the variability of task-induced DMN anatomic redistribution was hard to recapitulate, as none of the maps, or any linear combination of them, could represent the whole space of its dynamical reconfiguration. Altogether, our findings suggest that the DMN may be characterized by a richer functional diversity and a more spatial complexity than previously suggested.

neuroscience↗

bioRxiv2021

Direct tissue sensing reprograms TLR4+ Tfh-like cells inflammatory profile in the joints of rheumatoid arthritis patients

CD4+ T cells mediate rheumatoid arthritis (RA) pathogenesis through both antibody-dependent and independent mechanisms. It remains unclear how synovial microenvironment impinges on CD4+ T cells pathogenic functions. Here, we identified a TLR4+ follicular helper T (Tfh) cell-like population present in the blood and expanded in synovial fluid. Mechanistically, we unveiled that homotypic T-T cell interactions through non-cognate HLA-DR:TCR contacts regulate TLR4 expression on T cells. TLR4+ T cells possess a two-pronged pathogenic activity. Upon TCR and ICOS engagement, TLR4+ T cells produce IL-21, a cytokine known to sponsor antibody production. However, direct TLR4+ engagement on T cells, by endogenous ligands in the arthritic joint, reprograms them towards an IL-17 inflammatory profile compatible with tissue damage program. Blocking TLR4 signaling with a specific inhibitor impaired IL-17 production in response to synovial fluid recognition. Ex vivo, synovial fluid TLR4+ T cells produced IL-17, but not IL-21. TLR4+ T cells appear to uniquely reconcile an ability to promote systemic antibody production with a local synovial driven tissue damage program. TLR4+ T cells could constitute an attractive cellular target and predictive biomarker for erosive arthritis.

immunology↗

bioRxiv2020

OmpK36 and TraN facilitate conjugal transfer of the Klebsiella pneumoniae carbapenem resistance plasmid pKpQIL

We investigated the mechanism of conjugal transfer of the endemic Klebsiella pneumoniae carbapenem resistance plasmid, pKpQIL. Transfer efficiency of this plasmid was found to be dependent on the expression of the major outer membrane porin, OmpK36, in recipient cells. We also found that conjugal uptake is reduced in recipients expressing an OmpK36 isoform associated with the globally pervasive K. pneumoniae ST258 clade (OmpK36ST258). This reduction was attributed to a glycine-aspartate insertion in loop 3 of OmpK36ST258, which constricts the pore by 26%. Deletion of finO, which encodes an RNA-binding protein, derepressed transfer of pKpQIL and enabled visualisation of the conjugation pilus and OmpK36-dependent conjugation in real time. While deletion of traN abolished pKpQIL conjugation, substituting traN in pKpQIL with its homologue from R100-1 circumvented OmpK36 dependency. These results suggest that OmpK36 in recipient K. pneumoniae and the pKpQIL-encoded TraN in donor bacteria cooperate to facilitate plasmid transfer. This is the first report since 1998 to suggest a novel recipient cell receptor for IncF plasmid transfer and supports the idea that TraN mediates receptor specificity for plasmids belonging to this incompatibility group.

microbiology↗

bioRxiv2020

A meta-analytic approach to mapping co-occurrent grey matter volume increases and decreases in psychiatric disorders

Numerous studies have investigated gray matter (GM) volume changes in diverse patient groups. Reports of disorder-related GM reductions are common in such work, but many studies also report evidence for GM volume increases in patients. It is unclear whether these GM increases and decreases independent or related in some way. Here, we address this question using a novel meta-analytic network mapping approach. We used a coordinate-based meta-analysis of 64 voxel-based morphometry studies of psychiatric disorders to calculate the probability of finding a GM increase or decrease in one region given an observed change in the opposite direction in another region. Estimating this co-occurrence probability for every pair of brain regions allowed us to build a network of concurrent GM changes of opposing polarity. Our analysis revealed that disorder-related GM increases and decreases are not independent; instead, a GM change in one area is often statistically related to a change of opposite polarity in other areas, highlighting distributed yet coordinated changes in GM volume as a function of brain pathology. Most regions showing GM changes linked to an opposite change in a distal area were located in salience, executive-control and default mode networks, as well as the thalamus and basal ganglia. Moreover, pairs of regions showing coupled changes of opposite polarity were more likely to belong to different canonical networks than to the same one. Our results suggest that regional GM alterations in psychiatric disorders are often accompanied by opposing changes in distal regions that belong to distinct functional networks.

neuroscience↗

bioRxiv2019

Hubs of long-distance co-alteration in brain pathology

The exact mechanisms at the root of pathologic anatomical covariances are still unknown. It is nonetheless becoming clearer that the impact of brain diseases is more convincingly represented in terms of co-alterations rather than in terms of localization of alterations. According to this view, neurological and psychiatric conditions might be seen as whole-brain patterns of modifications. In this context, the physical distance between two co-altered areas may provide insightful information about how pathology develops across the brain, assuming that long-range co-alterations might be relevant features of pathological networks. To demonstrate this hypothesis, we calculated the probability of co-alteration between brain areas across a large database of voxel-based morphometry studies of psychiatric and neurological disorders, and we investigated the physical (Euclidean) distance of the edges of the resulting network. Such analysis produced a series of observations relevant for the understanding of pathology, which range from unanticipated results to the recognition of regions of well-known functional and clinical relevance. For instance, it emphasizes the importance of the anterior and dorsal prefrontal cortices in the distribution of the disease-related alterations, as well as a specular asymmetry of gray matter decreases and increases between the hemispheres. Also, the analyses of schizophrenia and Alzheimers disease show that long-distance co-alterations are able to identify areas involved in pathology and symptomatology. Moreover, the good concordance between the measure of the mean physical distance and that of functional degree centrality suggests that co-alterations and connectivity are intimately related. These findings highlight the importance of analyzing the physical distance in pathology, as the areas characterized by a long mean distance may be considered as hubs with a crucial role in the systems of alterations induced by brain diseases.

neuroscience↗

bioRxiv2019

Addressing reverse inference in structural brain alterations

In neuroimaging with reverse inferences we can infer the involvement of cognitive processes from certain patterns of brain activity. Still, the same reasoning holds if we substitute "brain activity" with "brain alteration" and "cognitive process" with "brain disorder". To assess the involvement of a certain alteration pattern in a brain disorder we used the Bayes factor technique on voxel-based morphometry data of schizophrenia and Alzheimers disease. This technique allows to calculate the ratio between the likelihoods of two alternative hypotheses (in our case, that the alteration of the voxel is specific for the brain disorder under scrutiny or that the alteration is not specific). We then performed temporal simulations of the alteration spread associated with different pathologies. The Bayes factor values calculated on these simulated data were able to reveal that the areas which are more specific to a certain disease are also the ones to be early altered.

neuroscience↗

bioRxiv2018

Weber’s law is the result of exact temporal accumulation of evidence

Webers law states that the discriminability between two stimulus intensities depends only on their ratio. Despite its status as the cornerstone of psychophysics, the mecha-nisms underlying Webers law are still debated, as no principled way exists to choose between its many proposed alternative explanations. We studied this problem training rats to discriminate the lateralization of sounds of different overall level. We found that the rats discrimination accuracy in this task is level-invariant, consistent with Webers law. Surprisingly, the shape of the reaction time distributions is also level-invariant, implying that the only behavioral effect of changes in the overall level of the sounds is a uniform scaling of time. Furthermore, we demonstrate that Webers law breaks down if the stimulus duration is capped at values shorter than the typical reaction time. Together, these facts suggest that Webers law is associated to a process of bounded evidence accumulation. Consistent with this hypothesis, we show that, among a broad class of sequential sampling models, the only robust mechanism consistent with reaction time scale-invariance is based on perfect accumulation of evidence up to a constant bound, Poisson-like statistics, and a power-law encoding of stimulus intensity. Fits of a minimal diffusion model with these characteristics describe the rats performance and reaction time distributions with virtually no error. Various manipulations of motivation were unable to alter the rats psychometric function, demonstrating the stability of the just-noticeable-difference and suggesting that, at least under some conditions, the bound for evidence accumulation can set a hard limit on discrimination accuracy. Our results establish the mechanistic foundation of the process of intensity discrimination and clarify the factors that limit the precision of sensory systems.

neuroscience↗

bioRxiv2017

Specific patterns of bold variability associated with the processing of pain stimuli

It is well known that the blood oxygen level dependent (BOLD) signal varies according to task performance and region specificity. This ongoing and fluctuating activity reflects the organization of functional brain networks. Peculiar dynamics of BOLD signal are therefore supposed to characterize brain activity in different conditions. Within this framework, we investigated through a multivoxel pattern analysis whether patterns of BOLD variability convey information that may allow an efficient discrimination between task (i.e., painful stimulation) and rest conditions. We therefore identified the most discriminative brain areas between the two conditions, which turned out to be the anterior insula, dorsal anterior cingulate cortex, posterior insula, the thalamus, and the periaqueductal gray. Then, on the basis of information theory, we calculated the entropy of their different time series. Entropy was found to distribute differently between these brain areas. The posterior insula was found to be is the smaller contributor to the entropy rate, whereas the system formed by the thalamus and periaqueductal gray was found to be the major contributor. Overall, the brain system reaches a higher level of entropy during the rest condition, which suggests that cerebral activity is characterized by a larger informational space when the brain is at rest than when it is engaged in a specific task. Thus, this study provides evidence that: i) the pattern of BOLD variance allow a good discrimination between the conditions of rest and pain stimulation; ii) the discriminative pattern resembles closely that of the functional network that has been called pain matrix; iii) brain areas with high and low variability are characterized by a different sample entropy; iv) the entropy rate of cerebral regions can be an insightful parameter to better understand the complex dynamics of the brain.

neuroscience↗

bioRxiv2017

Boredom begets creativity: a solution to the exploitation-exploration trade-off in predictive coding

Here, we investigate whether systems that minimize prediction error e.g. predictive coding, can also show creativity, or on the contrary, prediction error minimization unqualifies for the design of systems that respond in creative ways to non recurrent problems. We argue that there is a key ingredient that has been overlooked by researchers that needs to be incorporated to understand intelligent behavior in biological and technical systems. This ingredient is boredom. We propose a mathematical model based on the Black-Scholes-Merton equation which provides mechanistic insights into the interplay between boredom and prediction pleasure as the key drivers of behavior.

neuroscience↗