Integrating physics in deep learning algorithms: a force field as a PyTorch module.

MOTIVATION: Deep learning algorithms applied to structural biology often struggle to converge to meaningful solutions when limited data is available, since they are required to learn complex physical rules from examples. State-of-the-art force-fields, however, cannot interface with deep learning algorithms due to their implementation. RESULTS: We present MadraX, a forcefield implemented as a differentiable PyTorch module, able to interact with deep learning algorithms in an end-to-end fashion. AVAILABILITY AND IMPLEMENTATION: MadraX documentation, together with tutorials and installation guide, is available at madrax.readthedocs.io.

GeneGPT: augmenting large language models with domain tools for improved access to biomedical information.

MOTIVATION: While large language models (LLMs) have been successfully applied to various tasks, they still face challenges with hallucinations. Augmenting LLMs with domain-specific tools such as database utilities can facilitate easier and more precise access to specialized knowledge. In this article, we present GeneGPT, a novel method for teaching LLMs to use the Web APIs of the National Center for Biotechnology Information (NCBI) for answering genomics questions. Specifically, we prompt Codex to solve the GeneTuring tests with NCBI Web APIs by in-context learning and an augmented decoding algorithm that can detect and execute API calls. RESULTS: Experimental results show that GeneGPT achieves state-of-the-art performance on eight tasks in the GeneTuring benchmark with an average score of 0.83, largely surpassing retrieval-augmented LLMs such as the new Bing (0.44), biomedical LLMs such as BioMedLM (0.08) and BioGPT (0.04), as well as GPT-3 (0.16) and ChatGPT (0.12). Our further analyses suggest that: First, API demonstrations have good cross-task generalizability and are more useful than documentations for in-context learning; second, GeneGPT can generalize to longer chains of API calls and answer multi-hop questions in GeneHop, a novel dataset introduced in this work; finally, different types of errors are enriched in different tasks, providing valuable insights for future improvements. AVAILABILITY AND IMPLEMENTATION: The GeneGPT code and data are publicly available at https://github.com/ncbi/GeneGPT.

pyCapsid: identifying dominant dynamics and quasi-rigid mechanical units in protein shells.

SUMMARY: pyCapsid is a Python package developed to facilitate the characterization of the dynamics and quasi-rigid mechanical units of protein shells and other protein complexes. The package was developed in response to the rapid increase of high-resolution structures, particularly capsids of viruses, requiring multiscale biophysical analyses. Given a protein shell, pyCapsid generates the collective vibrations of its amino-acid residues, identifies quasi-rigid mechanical regions associated with the disassembly of the structure, and maps the results back to the input proteins for interpretation. pyCapsid summarizes the main results in a report that includes publication-quality figures. AVAILABILITY AND IMPLEMENTATION: pyCapsid's source code is available under MIT License on GitHub. It is compatible with Python 3.8-3.10 and has been deployed in two leading Python package-management systems, PIP and Conda. Installation instructions and tutorials are available in the online documentation and in the pyCapsid's YouTube playlist. In addition, a cloud-based implementation of pyCapsid is available as a Google Colab notebook. pyCapsid Colab does not require installation and generates the same report and outputs as the installable version. Users can post issues regarding pyCapsid in the repository's issues section.

KaMRaT: a C++ toolkit for k-mer count matrix dimension reduction.

MOTIVATION: KaMRaT is designed for processing large k-mer count tables derived from multi-sample, RNA-seq data. Its primary objective is to identify condition-specific or differentially expressed sequences, regardless of gene or transcript annotation. RESULTS: KaMRaT is implemented in C++. Major functions include scoring k-mers based on count statistics, merging overlapping k-mers into contigs and selecting k-mers based on their occurrence across specific samples. AVAILABILITY AND IMPLEMENTATION: Source code and documentation are available via https://github.com/Transipedia/KaMRaT.

EURISCO update 2023: the European Search Catalogue for Plant Genetic Resources, a pillar for documentation of genebank material.

The European Search Catalogue for Plant Genetic Resources (EURISCO) is a central entry point for information on crop plant germplasm accessions from institutions in Europe and beyond. In total, it provides data on more than two million accessions, making an important contribution to unlocking the vast genetic diversity that lies deposited in >400 germplasm collections in 43 countries. EURISCO serves as the reference system for the Plant Genetic Resources Strategy for Europe and represents a significant approach for documenting and making available the world's agrobiological diversity. EURISCO is well established as a resource in this field and forms the basis for a wide range of research projects. In this paper, we present current developments of EURISCO, which is accessible at http://eurisco.ecpgr.org.

Diagnosis and management of endometrial hyperplasia: A UK national audit of adherence to national guidance 2012-2020.

BACKGROUND: Endometrial hyperplasia (EH) is a precusor lesion for endometrial cancer (EC), the commonest gynaecological malignancy in high-income countries. EH is a proliferation of glandular tissue, classified as either non-atypical endometrial hyperplasia (NEH) or, if the cytological features are abnormal, atypical endometrial hyperplasia (AEH). The clinical significance of AEH is that patients face both a high risk of having occult EC and a high risk of progression to EC if untreated. Recommendations on the care of women with EH were introduced by United Kingdom-wide guidance (Green-top Guide No. 67, 2016). National adherence to guidance is unknown. We aimed to describe the care of patients with EH; to compare the patterns of care for those with EH with national guidance to identify opportunities for quality improvement; and to compare patterns of care prior to and following the introduction of national guidance to understand its impact. METHODS AND FINDINGS: In this UK-wide patient-level clinical audit, we included 3,307 women who received a new histological diagnosis of EH through a gynaecology service between 1 January 2012 and 30 June 2020. We described first-line management, management at 2 years, and surgical characteristics prior to and following national guidance for EH using proportions and 95% confidence intervals (CIs) and compared process measures between time periods using multilevel Poisson regression. Of the 3,307 patients, 1,570 had NEH and 1,511 had AEH between 2012 and 2019. An additional 85 patients had NEH and 141 had AEH during 2020. Prior to national guidance, 9% (95% CI [6%, 15%]) received no initial treatment for NEH compared with 3% (95% CI [1%, 5%]) post-guidance; 31% (95% CI [26%, 36%]) and 48% (95% CI [43% 53%]) received an intrauterine progestogen, respectively, in the same periods. The predominant management of women with AEH did not differ, with 68% (95% CI [61%, 74%]) and 67% (95 CI [63%, 71%]) receiving first-line hysterectomy, respectively. By 2 years, follow-up to histological regression without hysterectomy increased from 38% (95% CI [33%, 43%]) to 52% (95% CI [47%, 58%]) for those with NEH (rate ratio (RR) 1.38, 95% CI [1.18, 1.63] p < 0.001). We observed an increase in the use of total laparoscopic hysterectomy among those with AEH (RR 1.26, 95% CI [1.04, 1.52]). In the later period, 37% (95% CI [29%, 44%]) of women initially diagnosed with AEH who underwent a first-line hysterectomy, received an upgraded diagnosis of EC. Study limitations included retrospective data collection from routine clinical documentation and the inability to comprehensively understand the shared decision-making process where care differed from guidance. CONCLUSIONS: The care of patients with EH has changed in accordance with national guidance. More women received first-line medical management of NEH and were followed up to histological regression. The follow-up of those with AEH who do not undergo hysterectomy must be improved, given their very high risk of coexistent cancer and high risk of developing cancer.

GSEApy: a comprehensive package for performing gene set enrichment analysis in Python.

MOTIVATION: Gene set enrichment analysis (GSEA) is a commonly used algorithm for characterizing gene expression changes. However, the currently available tools used to perform GSEA have a limited ability to analyze large datasets, which is particularly problematic for the analysis of single-cell data. To overcome this limitation, we developed a GSEA package in Python (GSEApy), which could efficiently analyze large single-cell datasets. RESULTS: We present a package (GSEApy) that performs GSEA in either the command line or Python environment. GSEApy uses a Rust implementation to enable it to calculate the same enrichment statistic as GSEA for a collection of pathways. The Rust implementation of GSEApy is 3-fold faster than the Numpy version of GSEApy (v0.10.8) and uses >4-fold less memory. GSEApy also provides an interface between Python and Enrichr web services, as well as for BioMart. The Enrichr application programming interface enables GSEApy to perform over-representation analysis for an input gene list. Furthermore, GSEApy consists of several tools, each designed to facilitate a particular type of enrichment analysis. AVAILABILITY AND IMPLEMENTATION: The new GSEApy with Rust extension is deposited in PyPI: https://pypi.org/project/gseapy/. The GSEApy source code is freely available at https://github.com/zqfang/GSEApy. Also, the documentation website is available at https://gseapy.rtfd.io/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

ggkegg: analysis and visualization of KEGG data utilizing the grammar of graphics.

SUMMARY: The Kyoto Encyclopedia of Genes and Genomes (KEGG) database serves as a valuable systems biology resource and is widely utilized in diverse research fields. However, existing software does not allow flexible visualization and network analyses of the vast and complex KEGG data. We developed ggkegg, an R package that integrates KEGG information with ggplot2 and ggraph. ggkegg enables enhanced visualization and network analyses of KEGG data. We demonstrate the utility of the package by providing examples of its application in single-cell, bulk transcriptome, and microbiome analyses. ggkegg may empower researchers to analyze complex biological networks and present their results effectively. AVAILABILITY AND IMPLEMENTATION: The package and user documentation are available at: https://github.com/noriakis/ggkegg.

simpleaf: a simple, flexible, and scalable framework for single-cell data processing using alevin-fry.

SUMMARY: The alevin-fry ecosystem provides a robust and growing suite of programs for single-cell data processing. However, as new single-cell technologies are introduced, as the community continues to adjust best practices for data processing, and as the alevin-fry ecosystem itself expands and grows, it is becoming increasingly important to manage the complexity of alevin-fry's single-cell preprocessing workflows while retaining the performance and flexibility that make these tools enticing. We introduce simpleaf, a program that simplifies the processing of single-cell data using tools from the alevin-fry ecosystem, and adds new functionality and capabilities, while retaining the flexibility and performance of the underlying tools. AVAILABILITY AND IMPLEMENTATION: Simpleaf is written in Rust and released under a BSD 3-Clause license. It is freely available from its GitHub repository https://github.com/COMBINE-lab/simpleaf, and via bioconda. Documentation for simpleaf is available at https://simpleaf.readthedocs.io/en/latest/ and tutorials for simpleaf that have been developed can be accessed at https://combine-lab.github.io/alevin-fry-tutorials.

libSBOLj3: a graph-based library for design and data exchange in synthetic biology.

SUMMARY: The Synthetic Biology Open Language version 3 data standard provides a graph-based approach to exchange information about biological designs. The new data model has major updates and offers several features for software tools. Here, we present libSBOLj3 to facilitate data exchange and provide interoperability between computer-aided design and automation tools using this standard. The library adopts a graph-based approach. Tool developers can extend these graphs with application-specific information and use detailed validation reports to identify errors and interoperability issues and apply best practice rules. AVAILABILITY AND IMPLEMENTATION: The libSBOLj3 library is implemented in Java and can be downloaded or used as a Maven dependency. The open-source project, code examples and documentation about accessing and using the library are available via GitHub at https://github.com/SynBioDex/libSBOLj3.

Gonomics: uniting high performance and readability for genomics with Go.

SUMMARY: Many existing software libraries for genomics require researchers to pick between competing considerations: the performance of compiled languages and the accessibility of interpreted languages. Go, a modern compiled language, provides an opportunity to address this conflict. We introduce Gonomics, an open-source collection of command line programs and bioinformatic libraries implemented in Go that unites readability and performance for genomic analyses. Gonomics contains packages to read, write, and manipulate a wide array of file formats (e.g. FASTA, FASTQ, BED, BEDPE, SAM, BAM, and VCF), and can convert and interface between these formats. Furthermore, our modular library structure provides a flexible platform for researchers developing their own software tools to address specific questions. These commands can be combined and incorporated into complex pipelines to meet the growing need for high-performance bioinformatic resources. AVAILABILITY AND IMPLEMENTATION: Gonomics is implemented in the Go programming language. Source code, installation instructions, and documentation are freely available at https://github.com/vertgenlab/gonomics.

crosshap: R package for local haplotype visualization for trait association analysis.

SUMMARY: Genome-wide association studies (GWAS) excels at harnessing dense genomic variant datasets to identify candidate regions responsible for producing a given phenotype. However, GWAS and traditional fine-mapping methods do not provide insight into the complex local landscape of linkage that contains and has been shaped by the causal variant(s). Here, we present crosshap, an R package that performs robust density-based clustering of variants based on their linkage profiles to capture haplotype structures in a local genomic region of interest. Following this, crosshap is equipped with visualization tools for choosing optimal clustering parameters (ɛ) before producing an intuitive figure that provides an overview of the complex relationships between linked variants, haplotype combinations, phenotype, and metadata traits. AVAILABILITY AND IMPLEMENTATION: The crosshap package is freely available under the MIT license and can be downloaded directly from CRAN with R >4.0.0. The development version is available on GitHub alongside issue support (https://github.com/jacobimarsh/crosshap). Tutorial vignettes and documentation are available (https://jacobimarsh.github.io/crosshap/).

Cytoscape.js 2023 update: a graph theory library for visualization and analysis.

SUMMARY: Cytoscape.js is an open-source JavaScript-based graph library. Its most common use case is as a visualization software component, so it can be used to render interactive graphs in a web browser. It also can be used in a headless manner, useful for graph operations on a server, such as Node.js. This update describes new features and enhancements introduced over many new versions from 2015 to 2022. AVAILABILITY AND IMPLEMENTATION: Cytoscape.js is implemented in JavaScript. Documentation, downloads and source code are available at http://js.cytoscape.org. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

ASTER: accurately estimating the number of cell types in single-cell chromatin accessibility data.

SUMMARY: Recent innovations in single-cell chromatin accessibility sequencing (scCAS) have revolutionized the characterization of epigenomic heterogeneity. Estimation of the number of cell types is a crucial step for downstream analyses and biological implications. However, efforts to perform estimation specifically for scCAS data are limited. Here, we propose ASTER, an ensemble learning-based tool for accurately estimating the number of cell types in scCAS data. ASTER outperformed baseline methods in systematic evaluation on 27 datasets of various protocols, sizes, numbers of cell types, degrees of cell-type imbalance, cell states and qualities, providing valuable guidance for scCAS data analysis. AVAILABILITY AND IMPLEMENTATION: ASTER along with detailed documentation is freely accessible at https://aster.readthedocs.io/ under the MIT License. It can be seamlessly integrated into existing scCAS analysis workflows. The source code is available at https://github.com/biox-nku/aster. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Unbiased pangenome graphs.

MOTIVATION: Pangenome variation graphs model the mutual alignment of collections of DNA sequences. A set of pairwise alignments implies a variation graph, but there are no scalable methods to generate such a graph from these alignments. Existing related approaches depend on a single reference, a specific ordering of genomes or a de Bruijn model based on a fixed k-mer length. A scalable, self-contained method to build pangenome graphs without such limitations would be a key step in pangenome construction and manipulation pipelines. RESULTS: We design the seqwish algorithm, which builds a variation graph from a set of sequences and alignments between them. We first transform the alignment set into an implicit interval tree. To build up the variation graph, we query this tree-based representation of the alignments to reduce transitive matches into single DNA segments in a sequence graph. By recording the mapping from input sequence to output graph, we can trace the original paths through this graph, yielding a pangenome variation graph. We present an implementation that operates in external memory, using disk-backed data structures and lock-free parallel methods to drive the core graph induction step. We demonstrate that our method scales to very large graph induction problems by applying it to build pangenome graphs for several species. AVAILABILITY AND IMPLEMENTATION: seqwish is published as free software under the MIT open source license. Source code and documentation are available at https://github.com/ekg/seqwish. seqwish can be installed via Bioconda https://bioconda.github.io/recipes/seqwish/README.html or GNU Guix https://github.com/ekg/guix-genomics/blob/master/seqwish.scm.

Symbolic kinetic models in python (SKiMpy): intuitive modeling of large-scale biological kinetic models.

MOTIVATION: Large-scale kinetic models are an invaluable tool to understand the dynamic and adaptive responses of biological systems. The development and application of these models have been limited by the availability of computational tools to build and analyze large-scale models efficiently. The toolbox presented here provides the means to implement, parameterize and analyze large-scale kinetic models intuitively and efficiently. RESULTS: We present a Python package (SKiMpy) bridging this gap by implementing an efficient kinetic modeling toolbox for the semiautomatic generation and analysis of large-scale kinetic models for various biological domains such as signaling, gene expression and metabolism. Furthermore, we demonstrate how this toolbox is used to parameterize kinetic models around a steady-state reference efficiently. Finally, we show how SKiMpy can implement multispecies bioreactor simulations to assess biotechnological processes. AVAILABILITY AND IMPLEMENTATION: The software is available as a Python 3 package on GitHub: https://github.com/EPFL-LCSB/SKiMpy, along with adequate documentation. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

PyHMMER: a Python library binding to HMMER for efficient sequence analysis.

SUMMARY: PyHMMER provides Python integration of the popular profile Hidden Markov Model software HMMER via Cython bindings. This allows the annotation of protein sequences with profile HMMs and building new ones directly with Python. PyHMMER increases flexibility of use, allowing creating queries directly from Python code, launching searches, and obtaining results without I/O, or accessing previously unavailable statistics like uncorrected P-values. A new parallelization model greatly improves performance when running multithreaded searches, while producing the exact same results as HMMER. AVAILABILITY AND IMPLEMENTATION: PyHMMER supports all modern Python versions (Python 3.6+) and similar platforms as HMMER (x86 or PowerPC UNIX systems). Pre-compiled packages are released via PyPI (https://pypi.org/project/pyhmmer/) and Bioconda (https://anaconda.org/bioconda/pyhmmer). The PyHMMER source code is available under the terms of the open-source MIT licence and hosted on GitHub (https://github.com/althonos/pyhmmer); its documentation is available on ReadTheDocs (https://pyhmmer.readthedocs.io).

Cellenium-a scalable and interactive visual analytics app for exploring multimodal single-cell data.

SUMMARY: Multimodal single-cell sequencing data provide detailed views into the molecular biology of cells. To allow for interactive analyses of such rich data and to readily derive insights from it, new analysis solutions are required. In this work, we present Cellenium, our new scalable visual analytics web application that enables users to semantically integrate and organize all their single-cell RNA-, ATAC-, and CITE-sequencing studies. Users can then find relevant studies and analyze single-cell data within and across studies. An interactive cell annotation feature allows for adding user-defined cell types. AVAILABILITY AND IMPLEMENTATION: Source code and documentation are freely available under an MIT license and are available on GitHub (https://github.com/Bayer-Group/cellenium). The server backend is implemented in PostgreSQL, Python 3, and GraphQL, the frontend is written in ReactJS, TypeScript, and Mantine css, and plots are generated using plotlyjs, seaborn, vega-lite, and nivo.rocks. The application is dockerized and can be deployed and orchestrated on a standard workstation via docker-compose.

SIMBSIG: similarity search and clustering for biobank-scale data.

SUMMARY: In many modern bioinformatics applications, such as statistical genetics, or single-cell analysis, one frequently encounters datasets which are orders of magnitude too large for conventional in-memory analysis. To tackle this challenge, we introduce SIMBSIG (SIMmilarity Batched Search Integrated GPU), a highly scalable Python package which provides a scikit-learn-like interface for out-of-core, GPU-enabled similarity searches, principal component analysis and clustering. Due to the PyTorch backend, it is highly modular and particularly tailored to many data types with a particular focus on biobank data analysis. AVAILABILITY AND IMPLEMENTATION: SIMBSIG is freely available from PyPI and its source code and documentation can be found on GitHub (https://github.com/BorgwardtLab/simbsig) under a BSD-3 license.

screenwerk: a modular tool for the design and analysis of drug combination screens.

MOTIVATION: There is a rapidly growing interest in high-throughput drug combination screening to identify synergizing drug interactions for treatment of various maladies, such as cancer and infectious disease. This creates the need for pipelines that can be used to design such screens, perform quality control on the data and generate data files that can be analyzed by synergy-finding bioinformatics applications. RESULTS: screenwerk is an open-source, end-to-end modular tool available as an R-package for the design and analysis of drug combination screens. The tool allows for a customized build of pipelines through its modularity and provides a flexible approach to quality control and data analysis. screenwerk is adaptable to various experimental requirements with an emphasis on precision medicine. It can be coupled to other R packages, such as bayesynergy, to identify synergistic and antagonistic drug interactions in cell lines or patient samples. screenwerk is scalable and provides a complete solution for setting up drug sensitivity screens, read raw measurements and consolidate different datasets, perform various types of quality control and analyze, report and visualize the results of drug sensitivity screens. AVAILABILITY AND IMPLEMENTATION: The R-package and technical documentation is available at https://github.com/Enserink-lab/screenwerk; the R source code is publicly available at https://github.com/Enserink-lab/screenwerk under GNU General Public License v3.0; bayesynergy is accessible at https://github.com/ocbe-uio/bayesynergy. Selected modules are available through Galaxy, an open-source platform for FAIR data analysis at https://oncotools.elixir.no. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.