Biology subjects

Di Martino, E.

Publications and source records attributed to Di Martino, E..

6 recordsLinked to original sources

bioRxiv2024

BioEncoder: a metric learning toolkit for comparative organismal biology

In the realm of biological image analysis, deep learning (DL) has been a transformative force. Tasks like finding areas of interest within an image (segmentation) and/or sorting images into groups (classification), can now be automated and achieve unprecedented efficiency and accuracy. However, conventional DL methods are challenged by large-scale biodiversity datasets, which are characterized by unbalanced classes and small phenotypic differences between them. Here we present BioEncoder, a user-friendly toolkit for metric learning, which overcomes the aforementioned challenges by focussing on learning relationships between individual data points rather than on the separability of classes. This approach simplifies training of robust image classification models for biologists, democratizing access to advanced DL techniques. BioEncoder is encapsulated in a Python package, crafted for ease of use and flexibility across diverse datasets. It features taxon-agnostic data loaders, custom augmentation options, and streamlined hyperparameter adjustments through text-based configuration files. The toolkits significance lies in its potential to unlock new possibilities in biological image analysis - from phenomics and disease diagnosis to species identification and ecosystem monitoring. BioEncoder focuses on the urgent need for toolkits bridging the gap between complex DL methodologies and practical applications in biological research.

evolutionary biology↗

bioRxiv2022

DeepBryo: a web app for AI-assisted morphometric characterization of cheilostome bryozoans

O_LIBryozoans are becoming an increasingly popular study system in macroevolutionary, ecological, and paleobiological research. Members of this colonial invertebrate phylum are notable for displaying an exceptional degree of division of labor in the form of specialized modules (polymorphs), which allow for the inference of individual allocation of resources to reproduction, defense, and growth using simple morphometric tools. However, morphometric characterizations of bryozoans are notoriously labored, due to the high number of structures often captured per image, as well as the need for specialized knowledge necessary for classifying individual skeletal structures within those images. C_LIO_LIWe here introduce DeepBryo, a web application for deep learning-based morphometric characterization of cheilostome bryozoans. DeepBryo requires a single image as input and performs measurements automatically using instance segmentation algorithms. DeepBryo is capable of detecting objects belonging to six classes and outputting fourteen morphological shape measurements for each object based on the inferred segmentation maps. The users can visualize the predictions, check for errors, and directly filter model outputs on the web browser. Measurements can then be downloaded as a comma-separated values file. C_LIO_LIDeepBryo has been trained and validated on a total of 72,412 structures, belonging to six different object classes in 935 SEM images of cheilostome bryozoans belonging to 109 different families. The model shows high (>0.8) recall and precision for zooid-level structures. Its misclassification rate is low (~4%) and largely concentrated in a single object class (opesia). The models estimated structure-level area, height, and width measurements are statistically indistinguishable from those obtained via manual annotation (r2 varying from 0.89 to 0.98) and show no detectable bias. DeepBryo reduces the person-hours required for characterizing the zooids in individual colonies to less than 1% of the time required for manual annotation at no significant loss of measurement accuracy. C_LIO_LIOur results indicate that DeepBryo enables cost-, labor,- and time-efficient morphometric characterization of cheilostome bryozoans. DeepBryo can greatly increase the scale of macroevolutionary, ecological, taxonomic, and paleobiological analyses, as well as the accessibility of deep learning tools for this emerging model system. Finally, DeepBryo provides the building blocks necessary for adapting the current application to other study groups. C_LI

bioinformatics↗

bioRxiv2020

A broadly resolved molecular phylogeny of New Zealand cheilostome bryozoans as a framework for hypotheses of morphological evolution

Larger molecular phylogenies based on ever more genes are becoming commonplace with the advent of cheaper and more streamlined sequencing and bioinformatics pipelines. However, many groups of inconspicuous but no less evolutionarily or ecologically important marine invertebrates are still neglected in the quest for understanding species- and higher-level phylogenetic relationships. Here, we alleviate this issue by presenting the molecular sequences of 165 cheilostome bryozoan species from New Zealand waters. New Zealand is our geographic region of choice as its cheilostome fauna is taxonomically, functionally and ecologically diverse, and better characterized than many other such faunas in the world. Using this most taxonomically broadly-sampled and statistically-supported cheilostome phylogeny comprising 214 species, when including previously published sequences, we tested several existing systematic hypotheses based solely on morphological observations. We find that lower taxonomic level hypotheses (species and genera) are robust while our inferred trees did not reflect current higher-level systematics (family and above), illustrating a general need for the rethinking of current hypotheses. To illustrate the utility of our new phylogeny, we reconstruct the evolutionary history of frontal shields (i.e., a calcified bodywall layer in ascus-bearing cheilostomes) and asked if its presence has any bearing on the diversification rates of cheilostomes.

evolutionary biology↗

bioRxiv2020

Are local dominance and inter-clade dynamics causally linked when one fossil clade displaces another?

Disputing the supposition that ecological competition drives macroevolutionary patterns is now a familiar goal in many fossil biodiversity studies. But it is an elusive goal, hampered by patchy sampling, few assemblage-level comparative analyses, unverified ecological equivalence of clades and a dearth of appropriate statistical tools. We address these concerns with a fortified and vetted compilation of 40190 fossil species occurrences of cyclostome and cheilostome bryozoans, a canonical example of one taxonomically dominant clade being displaced by another. Dramatic increases in Cretaceous cheilostome genus diversification rates begin millions of years before cheilostomes overtake cyclostomes in local species proportions. Moreover, analyses of origination and extinction rates over 150 Myr suggest that inter-clade dynamics are causally linked to each other, but not to changing assemblage-level proportions. One Sentence SummaryGlobal fossil diversification rates and local taxonomic dominance are not causally linked.

paleontology↗

bioRxiv2020

Molecular phylogeny of historical micro-invertebrate specimens using de novo sequence assembly

Resolution of relationships at lower taxonomic levels is crucial for answering many evolutionary questions, and as such, sufficiently varied species representation is vital. This latter goal is not always achievable with relatively fresh samples. To alleviate the difficulties in procuring rarer taxa, we have seen increasing utilization of historical specimens in building molecular phylogenies using high throughput sequencing. This effort, however, has mainly focused on large-bodied or well-studied groups, with small-bodied and under-studied taxa under-prioritized. Here, we present a pipeline that utilizes both historical and contemporary specimens, to increase the resolution of phylogenetic relationships among understudied and small-bodied metazoans, namely, cheilostome bryozoans. In this study, we pioneer sequencing of air-dried bryozoans, utilizing a recent library preparation method for low DNA input. We use the de novo mitogenome assembly from the target specimen itself as reference for iterative mapping, and the comparison thereof. In doing so, we present mitochondrial and ribosomal RNA sequences of 43 cheilostomes representing 37 species, including 14 from historical samples ranging from 50 to 149 years old. The inferred phylogenetic relationships of these samples, analyzed together with publicly available sequence data, are shown in a statistically well-supported 65 taxa and 17 genes cheilostome tree. Finally, the methodological success is emphasized by circularizing a total of 27 mitogenomes, seven from historical cheilostome samples. Our study highlights the potential of utilizing DNA from micro-invertebrate specimens stored in natural history collections for resolving phylogenetic relationships between species.

evolutionary biology↗

bioRxiv2019

Lithium treatment reverses irradiation-induced changes in rodent neural progenitors

Cranial radiotherapy in children has detrimental effects on cognition, mood, and social competence in young cancer survivors. Treatments harnessing hippocampal neurogenesis are currently of great relevance in this context, and we previously showed that voluntary running introduced long after irradiation rescued hippocampal neurogenesis in young mice (Naylor et al. 2008a). Lithium, a well-known mood stabilizer, has both neuroprotective, pro-neurogenic as well as anti-tumor effects, and in the current study we introduced lithium treatment 4 weeks after irradiation, analogous to the voluntary running study. Female mice received a single 4 Gy whole-brain irradiation dose at postnatal day (PND) 21 and were randomized to 0.24% Li2CO3 chow or normal chow from PND 49 to 77. Hippocampal neurogenesis was assessed at PND 77, 91 and 105. We found that lithium treatment had a pro-proliferative effect on neural progenitors and promoted neuronal integration upon its discontinuation. Gene expression profiling and DNA methylation analysis identified two novel factors related to the observed effects, Tppp, associated with proliferation, and GAD2/65, associated with neuronal signaling. Our results show that lithium treatment reverses irradiation-induced impairment of hippocampal neurogenesis even when introduced long after the injury. We propose that lithium treatment should be intermittent in order to first make neural progenitors proliferate and then, upon discontinuation, allow them to differentiate. Our findings suggest that pharmacological treatment of cognitive so-called late effects in childhood cancer survivors is possible.

neuroscience↗