Clinical Applications of Machine Learning.

Objective: This review introduces interpretable predictive machine learning approaches, natural language processing, image recognition, and reinforcement learning methodologies to familiarize end users. Background: As machine learning, artificial intelligence, and generative artificial intelligence become increasingly utilized in clinical medicine, it is imperative that end users understand the underlying methodologies. Methods: This review describes publicly available datasets that can be used with interpretable predictive approaches, natural language processing, image recognition, and reinforcement learning models, outlines result interpretation, and provides references for in-depth information about each analytical framework. Results: This review introduces interpretable predictive machine learning models, natural language processing, image recognition, and reinforcement learning methodologies. Conclusions: Interpretable predictive machine learning models, natural language processing, image recognition, and reinforcement learning are core machine learning methodologies that underlie many of the artificial intelligence methodologies that will drive the future of clinical medicine and surgery. End users must be well versed in the strengths and weaknesses of these tools as they are applied to patient care now and in the future.

Phylogenomics of the Neotropical fish family Serrasalmidae with a novel intrafamilial classification (Teleostei: Characiformes).

The Neotropical fish family Serrasalmidae comprises 16 extant genera and 101 species widespread through major Neotropical rivers with relevant importance for regional fisheries and aquaculture. The monophyly of Serrasalmidae and the recognition of three main clades are recurrent between morphological and molecular phylogenies. However, both intergeneric and interspecific relationships within each of those clades remain uncertain. Here, we used 81 terminals of 69 species (68%) and all 16 genera of Serrasalmidae to sequence 1553 loci of ultraconserved elements (UCEs), multiple nuclear loci widely applied in phylogenetic studies, and performed maximum likelihood, Bayesian, and species tree analyses. We obtained highly supported phylogenies in all applied methods corroborating the monophyly of Serrasalmidae and the three-clade hypotheses herein proposed as two subfamilies and two tribes: (Colossomatinae (Serrasalminae (Myleini + Serrasalmini))). Morphological features for each subfamily involve the absence (Colossomatinae) or presence (Serrasalminae) of a pre-dorsal spine. Morphological diagnoses among tribes include the pre-dorsal spine being continuous (Myleini) or discontinuous (Serrasalmini) relative to the first unbranched dorsal-fin ray. Our results highlight the complexity of the relationships especially the non-monophyly of Myleus, Mylesinus, Myloplus, Tometes, and Utiaritichthys within Myleini, as well as of Serrasalmus and Pristobrycon within Serrasalmini.

Molecular delimitation and taxonomic revision of the wimple piranha Catoprion (Characiformes: Serrasalmidae) with the description of a new species.

A taxonomic revision of wimple piranhas of the genus Catoprion is performed in combination with a molecular analysis using mtDNA. Molecular phylogenetic analyses of 49 specimens using genetic distances, conventional likelihood and four delimitation methods yielded two distinct lineages of Catoprion, with the morphological analyses of 198 specimens of Catoprion corroborating the molecular results. We provide a redescription of Catoprion mento, from the Paraguay, Orinoco, and tributaries of western Amazon basin, keeping Mylesinus macropterus as a junior synonym of C. mento, and the description of Catoprion absconditus n. sp., from the Amazon and Essequibo basins. C. absconditus n. sp. differs from C. mento by the presence of 86-94 perforated scales in the lateral line (vs. 65-86 scales) and the presence of 35-40 circumpeduncular scales (vs. 29-34 scales). The distribution of C. mento follows the Amazonas-Paraguay-Orinoco lowlands, whereas C. absconditus follows the eastern Amazon biogeographic pattern.

Molecular Data Reveal Multiple Lineages in Piranhas of the Genus Pygocentrus (Teleostei, Characiformes).

Carnivorous piranhas are distributed in four serrasalmid genera including Pygocentrus, which inhabit major river basins of South America. While P. cariba and P. piraya are endemics of the Orinoco and São Francisco basins, respectively, P. nattereri is widely distributed across the Amazonas, Essequibo, lower Paraná, Paraguay, and coastal rivers of northeastern Brazil, with recent records of introductions in Asia. Few studies have focused on the genetic diversity and systematics of Pygocentrus and the putative presence of additional species within P. nattereri has never been the subject of a detailed molecular study. Here we aimed to delimit species of Pygocentrus, test the phylogeographic structure of P. nattereri, and access the origin of introduced specimens of P. nattereri in Asia. Phylogenetic analyses based on a mitochondrial dataset involving maximum-likelihood tree reconstruction, genetic distances, Bayesian analysis, three delimitation approaches, and haplotype analysis corroborate the morphological hypothesis of the occurrence of three species of Pygocentrus. However, we provide here strong evidence that P. nattereri contains at least five phylogeographically-structured lineages in the Amazonas, Guaporé (type locality), Itapecuru, Paraná/Paraguay, and Tocantins/Araguaia river basins. We finally found that the introduced specimens in Asia consistently descend from the lineage of P. nattereri from the main Rio Amazonas. These results contribute to future research aimed to detect morphological variation that may occur in those genetic lineages of Pygocentrus.

A multiple hypothesis approach to explain species richness patterns in neotropical stream-dweller fish communities.

Several hypotheses are used to explain species richness patterns. Some of them (e.g. species-area, species-energy, environment-energy, water-energy, terrestrial primary productivity, environmental spatial heterogeneity, and climatic heterogeneity) are known to explain species richness patterns of terrestrial organisms, especially when they are combined. For aquatic organisms, however, it is unclear if these hypotheses can be useful to explain for these purposes. Therefore, we used a selection model approach to assess the predictive capacity of such hypotheses, and to determine which of them (combined or not) would be the most appropriate to explain the fish species distribution in small Brazilian streams. We perform the Akaike's information criteria for models selections and the eigenvector analysis to control the special autocorrelation. The spatial structure was equal to 0.453, Moran's I, and require 11 spatial filters. All models were significant and had adjustments ranging from 0.370 to 0.416 with strong spatial component (ranging from 0.226 to 0.369) and low adjustments for environmental data (ranging from 0.001 to 0.119) We obtained two groups of hypothesis are able to explain the richness pattern (1) water-energy, temporal productivity-heterogeneity (AIC = 4498.800) and (2) water-energy, temporal productivity-heterogeneity and area (AIC = 4500.400). We conclude that the fish richness patterns in small Brazilian streams are better explained by a combination of Water-Energy + Productivity + Temporal Heterogeneity hypotheses and not by just one.

Taxonomic revision of the Cis-Andean species of Mylossoma Eigenmann Kennedy, 1903 (Teleostei: Characiformes: Serrasalmidae).

A revision of the cis-Andean species of Mylossoma is presented. Four species were recognized: M. albiscopum and M. aureum from the rio Amazonas and rio Orinoco basins; M. duriventre, from the rio Paraguai, lower rio Paraná and rio Uruguai basins; and M. unimaculatum, an endemic species from the Tocantins-Araguaia system. Both Mylossoma albiscopum and M. unimaculatum were removed from the synonymy of Mylossoma duriventre, which had its occurrence range restricted to the La Plata basin. The recognition of these four valid species corroborates the results of a previous DNA barcode study. Redescriptions of each species, and an identification key for the genus are provided.

ZZ/Z0: The New System of Sex Chromosomes in Eigenmannia aff. trilineata (Teleostei: Gymnotiformes: Sternopygidae) Characterized by Molecular Cytogenetics and DNA Barcoding.

The cytogenetic characteristics of Eigenmannia aff. trilineata were analyzed by basic and molecular cytogenetics, applying fluorescent in situ hybridization, with 18S and 5S rDNA and U2 snRNA probes. The species revealed a kind of polymorphism associated to ZZ/Z0 type sex chromosomes, with 2n = 32 (8m+2sm+22a, NF = 42) in all males under analysis, whereas females evidenced 2n = 31 (8m+1sm+22a, NF = 40). C-banding showed constitutive heterochromatin restricted to the pericentromeric region of all chromosomes and single-nucleolus organized regions on pair 11. A site for rDNA 5S was synthetic with a cluster of rDNA 18S near the centromere on the long arm of only one homologue of pair 11. Other clusters for 5S rDNA were sited on pairs 7, 10, 12, 13, and 16. Further, 5S rDNA was co-located with U2 cluster in the pericentromeric region of pair 12. Joint analysis of DNA barcoding from cytochrome c oxidase subunit I (COI) sequences, generated from the karyotyped samples of E. aff. trilineata, and sequences of other Gymnotiforms recognized E. aff. trilineata as an Operational Taxonomic Unit. Results foreground the hypothesis that cytotypes are independent evolution units as cryptic species with a low morphological differentiation level, although with high genetic/karyotype differentiation rates.

Molecular identification of cryptic diversity in species of cis-Andean Mylossoma (Characiformes: Serrasalmidae).

Mylossoma is a Serrasalmidae genus with only two current valid species in the cis-Andean region but with several available names, today considered as junior synonymous. Morphological information combined with single-locus DNA sequences of cytochrome c oxidase I gene analysed by Barcode Index Number and General Mixed Yule Coalescent model were used in the present study to help the recognition of Operational Taxonomic Units (OTUs) in cis-Andean Mylossoma and discuss species boundaries within the genus. Five OTUs were recognized based on both morphological and molecular approaches. The analysis using the Barcode Index Number resulted in five OTUs, with M. duriventre being split in one unity in the Amazon, one in the Orinoco, one in Paraná-Paraguay and one in Tocantins-Araguaia which is coherent with our morphological results.