Metric Multidimensional Scaling for Large Single-Cell Data Sets using Neural Networks

Metric multidimensional scaling is one of the classical methods for embedding data into low-dimensional Euclidean space. It creates the low-dimensional embedding by approximately preserving the pairwise distances between the input points. However, current state-of-the-art approaches only scale to a few thousand data points. For larger data sets such as those occurring in single-cell RNA sequencing experiments, the running time becomes prohibitively large and thus alternative methods such as PCA are widely used instead. Here, we propose a neural network based approach for solving the metric multidimensional scaling problem that is orders of magnitude faster than previous state-of-the-art approaches, and hence scales to data sets with up to a few million cells. At the same time, it provides a non-linear mapping between high- and low-dimensional space that can place previously unseen cells in the same embedding.

Morfologi dan variasi morfometrik stingless bees di Kepulauan Maluku, Indonesia

Lebah tanpa sengat tersebar luas di wilayah tropis termasuk Indonesia dan tercatat tiga spesies lebah tanpa sengat di Maluku. Penelitian ini bertujuan untuk mengeksplorasi spesies lebah tanpa sengat dan persebarannya di Kepulauan Maluku, yaitu di Pulau Seram, Ambon, Haruku, Saparua, dan Nusalaut, serta membandingkan variasi morfometrik lebah tanpa sengat antar spesies dan antar individu dari spesies yang sama pada setiap pulau serta antar pulau. Sampel lebah diidentifikasi berdasarkan morfologi dengan 12 parameter morfometrik. Penelitian ini berhasil mengidentifikasi dua spesies lebah tanpa sengat, yaitu Tetragonula fuscobalteata (Cameron) dan T. sapiens (Cockerell). T. fuscobalteata merupakan catatan baru lebah tanpa sengat di Kepulauan Maluku dan lebah ini secara nyata memiliki pengukuran morfometrik lebih rendah daripada T. sapiens (P < 0,001). Berdasarkan analisis ordinasi non-metric multidimensional scaling (NMDS), gabungan 12 parameter morfometrik antara T. fuscobalteata dan T. sapiens menunjukkan perbedaan yang nyata. Parameter morfometrik T. fuscobalteata sangat bervariasi antar pulau kecuali satu parameter morfometrik (lebar gena) yang tidak nyata antar pulau (P > 0,05). Hal menarik lain berdasarkan analisis NMDS, koloni T. fuscobalteata di Ambon, Haruku, Saparua, dan Nusalaut sebagian besar membentuk kelompok, namun di Pulau Seram lebah tersebar dengan variasi morfometrik yang tinggi. T. sapiens ditemukan di tiga pulau, yaitu Seram, Ambon, dan Haruku. Lebar toraks, mesonotum dan panjang propodeum T. sapiens berbeda nyata di antara tiga pulau (P < 0,001). Hasil penelitian ini berkontribusi dalam menambah catatan penemuan persebaran baru T. fuscobalteata dan T. sapiens di Kepulauan Maluku dan menunjukkan variasi parameter morfometrik pada spesies lebah tanpa sengat antar pulau di Maluku.

Bat (Mammalia: Chiroptera) biodiversity in a subtropical inselberg ecosystem of Northeastern Argentina

Introduction: Rocky outcrops influence both micro and macro habitat for plants and animals by increasing the availability and predictability of food, providing shelter, and unique microclimatic conditions. Objective: We describe the bat assemblage in three isolated rocky outcrops and their surrounding pediments, adding new data about trophic guilds and roosts. Methods: We surveyed the bat fauna of Paraje Tres Cerros, Corrientes Province, Argentina, explored 13 sites during 26 field trips, using mist nets and search of roosts. We made an accumulation species curve and calculate inventory completeness. The similarity and composition of different habitats was compared by a non-metric multidimensional scaling (NMDS) and an ANOSIM test. Results: We reported 13 species, with a completeness inventory of the 80 % (Jacknife1), corresponding to the families Vespertilionidae, Molossidae and Phyllostomidae. The insectivorous bats were the best represented guild. We found nine types of roost for eight species in different sites. The analysis expressed low differences in the species composition between both types of habitats in the area. Conclusions: The rocky outcrops and their pediments are important ecosystems for the conservation and diversity of the bats in the region, since their communities are regionally unique.

Montane mire vegetation of the New England Tablelands Bioregion of Eastern Australia

Aims: To use unsupervised techniques to produce a hierarchical classification of montane mires of the study region. Study area: New England Tablelands Bioregion (NETB) of eastern Australia. Methods: A dataset of 280 vascular floristic survey plots placed across the variation in montane mires of the NETB was collated. Vegetation types were identified with the aid of a clustering method based on group averaging and tested using similarity profile analysis (SIMPROF) and through ordinations using Bray-Curtis similarity and non-metric multidimensional scaling (NMDS). A hierarchical schema was developed based on EcoVeg hierarchy and was circumscribed using positive and negative diagnostic taxa via similarity percentage analysis (SIMPER) and importance based on summed cover scores and frequency. Results: We defined one macrogroup to include all montane mire vegetation of the NETB and within these two groups and twelve alliances. Conclusions: Our study re-enforced the separation of bogs from other montane mire systems and confirmed the separation of fens and wet meadows, a distinction that previously had not been independently tested. Based on our results many existing montane mire communities of the NETB have been ill-defined at multiple hierarchical levels, leading to confusion in threat status and mapping. Additionally, nearly half of the alliances we recognise were found to have no correlates within current classification systems, which necessarily has implications for the effectiveness of current conservation planning. Taxonomic reference: PlantNET (http://plantnet.rbgsyd.nsw.gov.au/, accessed June 2016). Abbreviations: BC Act = Biodiversity Conservation Act; EPBC Act = Environmental Protection and Biodiversity Act; NETB = New England Tablelands Bioregion; NMDS = non-metric multidimensional scaling; PCT = plant community type; RE = regional ecosystem; SIMPER = similarity percentage analysis; SIMPROF = similarity profile analysis.

Finite Mixture Model-based classification of a complex vegetation system

Aim: To propose a Finite Mixture Model (FMM) as an additional approach for classifying large datasets of georeferenced vegetation plots from complex vegetation systems. Study area: The Italian peninsula including the two main islands (Sicily and Sardinia), but excluding the Alps and the Po plain. Methods: We used a database of 5,593 georeferenced plots and 1,586 vascular species of forest vegetation, created in TURBOVEG by storing published and unpublished phytosociological plots collected over the last 30 years. The plots were classified according to species composition and environmental variables using a FMM. Classification results were compared with those obtained by TWINSPAN algorithm. Groups were characterized in terms of ecological parameters, dominant and diagnostic species using the fidelity coefficient. Interpretation of resulting forest vegetation types was supported by a predictive map, produced using discriminant functions on environmental predictors, and by a non‐metric multidimensional scaling ordination. Results: FMM clustering obtained 24 groups that were compared with those from TWINSPAN, and similarities were found only at a higher classification level corresponding to the main orders of the Italian broadleaf forest vegetation: Fagetalia sylvaticae, Carpinetalia betuli, Quercetalia pubescenti-petraeae and Quercetalia ilicis. At lower syntaxonomic level, these 24 groups were referred to alliances and sub-alliances. Conclusions: Despite a greater computational complexity, FMM appears to be an effective alternative to the traditional classification methods through the incorporation of modelling in the classificatory process. This allows classification of both the co-occurrence of species and environmental factors so that groups are identified not only on their species composition, as in the case of TWINSPAN, but also on their specific environmental niche. Taxonomic reference: Conti et al. (2005). Abbreviations: CLM = Community-level models; FMM = Finite Mixture Model; NMDS = non‐metric multidimensional scaling.

Phenotypic analysis of aposematic conoderine weevils (Coleoptera: Curculionidae: Conoderinae) supports the existence of three large mimicry complexes

The Conoderinae (Coleoptera: Curculionidae) are one of the most distinctive Neotropical weevil groups in behaviour and appearance, attracting numerous hypotheses regarding the evolution and function of widespread apparent mimetic convergence. Conoderines have a poorly documented natural history, and a large fraction of the diversity of the group remains undescribed, presenting challenges to their study. In this analysis, 128 species of conoderine weevils previously or herein hypothesized to belong to three mimicry complexes are analysed in the first quantitative test of conoderine mimicry. Fifteen continuous and categorical characters describing the size, shape and coloration of these weevils were analysed using non-metric multidimensional scaling while statistically testing the resulting clusters in ordination space. Three similar, putatively mimetic complexes are recognized: (1) the ‘red-eyed fly’ complex of weevils, which are hypothesized to be evasively mimetic on various species of red-eyed flies; (2) the ‘striped/spotted’ complex, composed of weevils with a brightly coloured pronotum and red to white elytral stripes or spots; and (3) the ‘shiny blue’ complex of species with iridescent blue to blue–green pronotal scales. Each of these groups covers a wide geographical distribution and has evolved independently in multiple genera, although the red-eyed fly complex appears to be both the most species rich and widely distributed phylogenetically. Groupings were found to be statistically significant, although variation within each group suggests that the similarity in appearance of species in each group could be attributable to independent convergence on different, but phenotypically similar, models. Several avenues for future research on conoderine mimicry are discussed.