Biology subjects

Porto, A.

Publications and source records attributed to Porto, A..

12 recordsLinked to original sources

bioRxiv2026

Biology-guided machine learning for robust anatomical correspondence across biodiversity datasets

Three-dimensional imaging is rapidly increasing the scale of biological morphometric datasets, but establishing automated anatomical correspondence remains a major bottleneck. Conventional point-cloud registration methods transfer landmarks efficiently, yet poor initialization or high morphological variation can produce geometrically plausible but biologically implausible deformations. Here, we developed a statistical-shape-model (SSM)-regularized coherent point drift (CPD) estimator for automated morphometric landmark transfer. It replaces the unconstrained non-rigid CPD update with optimization in SSM coefficient space and regularizes coefficients by inverse population covariance, so deformations along low-variance biological modes are penalized more strongly. We compared pose-matched conventional CPD, SSM-CPD, and SSM-CPD with refinement in leave-one-out atlas experiments on laboratory mouse skulls (N=249), primate molars (N=115), and squamate vertebrae (N=254) under aligned, mildly perturbed, and severely perturbed initialization. We evaluated these formulations in terms of landmark accuracy, failure-tail behavior, runtime, and preservation of morphometric structure. Under severe perturbation, we also evaluated an integrated recovery procedure that optimizes the starting template, recovers global pose, and then applies SSM-CPD with refinement. Severe misalignment produced widespread failures in pose-matched conventional CPD, whereas SSM regularization reduced median error from 13 - 21% to approximately 1 - 3% of the target bounding-box diagonal. This improvement was highly consistent for skulls and molars. In the morphologically complex vertebral dataset, however, low median error concealed a substantial minority of catastrophic failures (28% of vertebrae above a 6% error threshold), which integrated recovery reduced to 4%. Across datasets, automated landmarks preserved the major axes and pairwise relationships of morphometric shape space, and local CPD refinement improved covariance preservation. Embedding learned biological variation directly into registration broadens the range of starting conditions and anatomies for which automated landmark transfer remains reliable. Our results also show that failure tails and downstream morphometric fidelity must be assessed alongside median accuracy when automated landmarking methods are evaluated.

bioinformatics↗

bioRxiv2026

LizardMorph: A generalizable machine learning framework for automated anatomical landmark detection in digital images

Morphological measurements underpin a wide range of ecological and evolutionary research, yet the manual landmarking workflows on which most morphometric studies depend remain a persistent bottleneck that limits both the pace and scale of biological research. Machine learning offers compelling solutions, but most automated landmarking tools require substantial computational expertise, creating a gap between technical capability and practical adoption by biologists. Here, we present LizardMorph, an integrated machine learning pipeline and web-based interface for semi-automated anatomical landmark detection on biological images. LizardMorph couples a fine-tuned ML-Morph shape predictor with an accessible, browser-based interface that enables researchers to upload images, review automated landmark predictions, interactively correct outliers through point-and-click editing, and export results in standard morphometric formats--all without programming expertise or local software installation. Using dorsal X-ray radiographs of Anolis lizards with 34 anatomical landmarks as a proof-of-concept, we show that the ML-Morph model achieves high predictive accuracy, with landmarks on well-defined skeletal structures predicted with 100% accuracy within a 1 mm tolerance threshold. A controlled user study comparing LizardMorph against traditional manual landmarking (TpsDig2) demonstrated significant efficiency gains: experienced annotators completed LizardMorph landmark verification 37.5% faster than manual annotation. Extrapolated to batch processing 1,000 lizards, LizardMorph saves experienced researchers approximately 6.5 hours of manual processing time. Critically, LizardMorph implements a human-in-the-loop design in which automated predictions serve as editable starting points, preserving researcher oversight and enabling correction of the occasional large-error outliers that would be unacceptable in fully automated workflows. LizardMorph is freely available as an open-source tool and provides a replicable framework for developing ML-assisted annotation tools that can democratize access to high-quality morphometric analysis across diverse biological research communities.

evolutionary biology↗

bioRxiv2025

Invasion of new adaptive zones retains telltale signs of directional selection at macroevolutionary scales in mammals

Directional selection is often viewed as a transient force in macroevolution, with its signal eroded over time by stabilizing and fluctuating selection. Yet, transitions into new adaptive zones are predicted to impose strong and sustained selective pressures that may leave a detectable signature even across deep timescales. We test this prediction by comparing the rates of multivariate skull morphological evolution required to traverse the boundaries between adaptive zones against genetic drift expectations. Our dataset includes 11,793 specimens spanning 231 species from 12 mammalian clades, each containing unique ecological transitions into new adaptive zones. Using a quantitative genetics framework, we estimated the phenotypic distances between ancestral and derived adaptive zones and contrasted them with null expectations under genetic drift. While a few adaptive zone invasions (e.g., marsupials and rodents) are consistent with drift, most exhibit substantially elevated rates of evolution. These results suggest that directional selection has recurrently shaped mammalian cranial evolution during major ecological shifts. We propose that adaptive zone transitions represent evolutionary contexts in which adaptation leaves a persistent macroevolutionary signal, challenging the prevailing view that long-term patterns are dominated by static forces.

evolutionary biology↗

bioRxiv2025

Aposematic color patterns are the dominant axis of phenotypic diversification in Nymphalid butterflies

Butterfly wing patterns serve diverse roles in visual communication, from aposematic signaling and mimicry to mate attraction and camouflage. In brush-footed butterflies (Nymphalidae), this diversity can be traced to the wing pattern "ground plan" that generates phenotypes deterministic in origin yet highly multidimensional in form. Quantifying such complexity at scale has long been a challenge, limiting our understanding of how visual signals interact, constrain one another, and evolve. Here, we used computer vision to extract high-dimensional traits from standardized museum specimens and assembled the largest comparative dataset of wing color patterns to date, spanning over one third of all known Nymphalid species. We first tested whether chemically defended species occupy a distinct region in morphospace and then derived a quantitative score for aposematism from the principal color patterns associated with defense. Using this score, we examined whether aposematic signals are expressed consistently across wing surfaces and sexes, and whether their origins are linked to shifts in evolutionary rate. We found that the dominant axis of morphospace is defined by chromatic and achromatic contrast, along which defended and undefended species cluster. Validation with an expert-labeled moth dataset confirmed that this axis separates aposematic from non-aposematic phenotypes across Lepidoptera. Consistent with theory, strongly aposematic species showed greater visual similarity between dorsal and ventral surfaces, between sexes, and among individuals. Rate analyses further indicated that aposematic patterns evolved repeatedly and were associated with non-linear shifts in evolutionary tempo. Together, these results identify aposematism as the dominant organizing axis of wing color pattern evolution in Nymphalidae. Significance statementButterflies are renowned for their striking diversity of wing patterns, including the warning colors that signal chemical defense to predators. Yet whether such warning patterns share common features across lineages has remained unclear. Here, we applied a metric computer vision model to more than 16,000 museum specimens spanning one third of all Nymphalid species, encoding their wing patterns into a common morphospace. Within this space, high-contrast aposematic patterns emerged as the dominant axis of diversification, explaining up to 20% of phenotypic variance. These signals were expressed consistently across wing surfaces, sexes, and individuals, and evolved repeatedly across the family. Our approach demonstrates how computer vision enables meaningful comparative analyses of complex patterns, revealing general principles of butterfly diversification.

evolutionary biology↗

bioRxiv2025

Appendometer: A system for simultaneous, high-throughput morphometry of Drosophila legs and wings

O_LIThe inheritance of many different organismal features is correlated, as is their evolution, sug-gesting that we need to understand the pattern and causes of those correlations to understand evolution. Unfortunately, we generally lack the ability to rapidly and accurately measure large numbers of traits, making it difficult to describe the patterns of trait relationships or for-mulate hypothesis about the causes of their entanglement. We have previously developed a system to make high-dimensional measurements of Drosophilid fly wings in live specimens. Here, we report the extension of this approach to rapidly assess the dimensions of the distal leg segments, femur, tibia, and tarsi. Using the system, we describe the covariance of the wing and leg morphology and evaluate the relative rates of evolution of legs and wings. C_LIO_LIWe use two simple suction devices to immobilize and position legs and wings of an anaes-thetized fly for imaging, then take a single image incorporating both appendages. We em-ployed a machine learning method to measure leg segment lengths, which should be broadly applicable across diverse taxa. Experienced users can image the legs and wings of a fly every two minutes, with outlier detection and correction taking approximately 40 seconds. To demonstrate the usefulness of these methods, we measured the legs and wings of over 4,000 specimens from 43 different Drosophilid taxa. We estimated the rate of wing and leg evolu-tion using a phylogenetic mixed model. C_LIO_LIRepeatabilities of leg segments lengths averaged over 80%. The rate of evolution of wing and leg sizes are similar, but the rate of wing proportion evolution is 1.6 times as high as that of leg proportions due to strong allometric changes in wing shape. Within-species variation in leg proportions is highly correlated with the rate of leg proportion evolution, as is true for wings. Relative lengths of leg segments showed a strong pattern of negative correlations be-tween the lengths of the tarsal segments and of the femurs and tibias, while all other segment correlations were positive. This pattern was repeated in the rate estimates, suggesting that se-lection favors tradeoffs between tarsi and the remainder of the leg. C_LIO_LIOur simple system for imaging and measuring legs and wings simultaneously has high throughput and repeatability. It is readily applicable to a wide variety of winged insects and other traits, including wings, that could be imaged. Applied to Drosophila, our morphometric system enlarges the ability to study inheritance, pleiotropy, and evolution in this important model taxon. C_LI

evolutionary biology↗

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↗

bioRxiv2022

Automated Landmarking via Multiple Templates

Geometric morphometrics based on landmark data has been increasingly used in biomedical and biological research for quantifying complex phenotypes. However, manual landmarking can be laborious and subject to intra and interobserver errors. This has motivated the development of automated landmarking methods. We have recently introduced ALPACA (Automated Landmarking through Point Cloud Alignment and Correspondence), a fast method to automatically annotate landmarks via the use of a landmark template as a part of the SlicerMorph toolkit. Yet, using a single template may not consistently perform well for large study samples, especially when the sample consists of specimens with highly variable morphology, as it is common in evolutionary studies. In this study, we introduce a variation on our ALPACA pipeline that supports multiple specimen templates, which we call MALPACA. We show that MALPACA outperforms ALPACA consistently by testing on two different datasets. We also introduce a method of choosing the templates that can be used in conjunction with MALPACA, when no prior information is available. This K-means method uses an approximation of the total morphological variation in the dataset to suggest samples within the population to be used as landmark templates. While we advise investigators to pay careful attention to the template selection process in any of the template-based automated landmarking approaches, our analyses show that the introduced K-means based method of templates selection is better than randomly choosing the templates. In summary, MALPACA can accommodate larger morphological disparity commonly found in evolutionary studies with performance comparable to human observer.

evolutionary biology↗

bioRxiv2021

Morphological integration during postnatal ontogeny: implications for evolutionary biology

How covariance patterns of phenotypes change during development is fundamental for a broader understanding of evolution. There is compelling evidence that mammalian skull covariance patterns change during ontogeny. However, it is unclear to what extent variation in covariance patterns during ontogeny can impact the response to selection. To tackle this question we explored: i) the extent to which covariance patterns change during postnatal ontogeny; ii) in which ontogenetic stages covariance patterns differ the most, and iii) the extent to which the phenotypic covariance pattern at different ontogenetic stages can be explained by the same processes determining additive genetic covariance. We sampled postnatal ontogenetic series for both marsupials, and placentals. Within each ontogenetic series, we compared covariance matrices (P-matrices) at different ontogenetic stages. Furthermore, we compared these P-matrices to two target matrices [adult P-matrix and an additive genetic covariance matrix (G-matrix)]. Our results show that for all ontogenetic series, covariance patterns from weaning onward are conserved and probably shaped by the same processes determining the G-matrix. We conclude that irrespective of eventual differences in how selection operates during most of postnatal ontogeny, the net response to such pressures will probably not be affected by ontogenetic differences in the covariance pattern.

developmental biology↗

bioRxiv2020

SlicerMorph: An open and extensible platform to retrieve, visualize and analyze 3D morphology

Large scale digitization projects such as #ScanAllFishes and oVert are generating high-resolution microCT scans of vertebrates by the thousands. Data from these projects are shared with the community using aggregate 3D specimen repositories like MorphoSource through various open licenses. MorphoSource currently hosts tens of thousands of 3D scans of eukaryotes. Along with the data from similarly scoped projects such as 10kPhenomes, DigiMorph and many others, soon hundreds of thousands of specimens that represent biodiversity of extinct and extant organisms will be conveniently available to researchers. We anticipate an explosion of quantitative research in organismal biology with the convergence of available data and the methodologies to analyze them. Though the data are available, the road from a series of images to analysis is fraught with challenges for most biologists. It involves tedious tasks of data format conversions, preserving spatial scale of the data accurately, 3D visualization and segmentations, acquiring measurements and annotations. When scientists use commercial software with proprietary formats, a roadblock for data exchange, collaboration, and reproducibility is erected that hurts the efforts of the scientific community to broaden participation in research. Another relevant concern is that ultimate derivative data from individual research projects (e.g., 3D models of segmentation) are shared in formats that do not preserve the correct spatial scale of the data. In this paper, we present our effort to tackle challenges biologists face when conducting 3D specimen-based research. We developed SlicerMorph as an extension of 3D Slicer, a biomedical visualization and analysis ecosystem with extensive visualization and segmentation capabilities built on proven python-scriptable open-source libraries such as Visualization Toolkit and Insight Toolkit. In addition to the core functionalities of Slicer, SlicerMorph provides users with modules to conveniently retrieve open-access 3D models or import users own 3D volumes, to annotate 3D curve and patch-based landmarks, generate canonical templates, conduct geometric morphometric analyses of 3D organismal form using both landmark-driven and landmark-free approaches, and create 3D animations from their results. We highlight how these individual modules can be tied together to establish complete workflow(s) from image sequence to morphospace. Our software development efforts were supplemented with short courses and workshops that cover the fundamentals of 3D imaging and morphometric analyses as it applies to study of organismal form and shape in evolutionary biology, and extensive links to the existing tutorials are provided as supplemental material. Our goal is to establish a community of organismal biologists centered around Slicer and SlicerMorph to facilitate easy exchange of data and results and collaborations using 3D specimens. Our proposition to our colleagues is that using a common open platform supported by a large user and developer community ensures the longevity and sustainability of the tools beyond the initial development effort.

evolutionary biology↗

bioRxiv2020

ALPACA: a fast and accurate approach for automated landmarking of three-dimensional biological structures

Landmark-based geometric morphometrics has emerged as an essential discipline for the quantitative analysis of size and shape in ecology and evolution. With the ever-increasing density of digitized landmarks, the possible development of a fully automated method of landmark placement has attracted considerable attention. Despite the recent progress in image registration techniques, which could provide a pathway to automation, three-dimensional morphometric data is still mainly gathered by trained experts. For the most part, the large infrastructure requirements necessary to perform image-based registration, together with its system-specificity and its overall speed have prevented wide dissemination. Here, we propose and implement a general and lightweight point cloud-based approach to automatically collect highdimensional landmark data in 3D surfaces (Automated Landmarking through Point cloud Alignment and Correspondence Analysis). Our framework possesses several advantages compared with image-based approaches. First, it presents comparable landmarking accuracy, despite relying on a single, random reference specimen and much sparser sampling of the structures surface. Second, it is performant such that it can be efficiently run on consumer-grade personal computers. Finally, it is general and can be applied to any biological structure of interest, regardless of whether anatomical atlases are available. Our validation procedures indicate that the method is capable of recovering multivariate patterns of morphological variation that are largely indistinguishable from those obtained by manual digitization, indicating that the use of an automated landmarking approach should not result in different conclusions regarding the nature of multivariate patterns of morphological variation. The proposed point cloud-based approach has the potential to increase the scale and reproducibility of morphometrics research. To allow ALPACA to be used out-of-the-box by users with no prior programming experience, we implemented it as a module as part of the SlicerMorph project. SlicerMorph is an extension that enables geometric morphometrics data collection and 3D specimen analysis within the open-source 3D Slicer biomedical visualization ecosystem. We expect that convenient access to this platform will make ALPACA broadly applicable within ecology and evolution.

evolutionary biology↗

bioRxiv2019

ML-morph: A fast, accurate and general approach for automated detection and landmarking of biological structures in images

O_LIMorphometrics has become an indispensable component of the statistical analysis of size and shape variation in biological structures. Morphometric data has traditionally been gathered through low-throughput manual landmark annotation, which represents a significant bottleneck for morphometric-based phenomics. Here we propose a machine-learning-based high-throughput pipeline to collect high-dimensional morphometric data in images of semi rigid biological structures.\nC_LIO_LIThe proposed framework has four main strengths. First, it allows for dense phenotyping with minimal impact on specimens. Second, it presents landmarking accuracy comparable to manual annotators, when applied to standardized datasets. Third, it performs data collection at speeds several orders of magnitude higher than manual annotators. And finally, it is of general applicability (i.e., not tied to a specific study system).\nC_LIO_LIState-of-the-art validation procedures show that the method achieves low error levels when applied to three morphometric datasets of increasing complexity, with error varying from 0.5% to 2% of the structures length in the automated placement of landmarks. As a benchmark for the speed of the entire automated landmarking pipeline, our framework places 23 landmarks on 13,686 objects (zooids) detected in 1684 pictures of fossil bryozoans in 3.12 minutes using a personal computer.\nC_LIO_LIThe proposed machine-learning-based phenotyping pipeline can greatly increase the scale, reproducibility and speed of data collection within biological research. To aid the use of the framework, we have developed a file conversion algorithm that can be used to leverage current morphometric datasets for automation, allowing the entire procedure, from model training all the way to prediction, to be performed in a matter of hours.\nC_LI

evolutionary biology↗