Find open-source science resources

This packages simulates spatial transcriptomics data with the mean- variance relationship using a Gaussian Process model per gene.

> [!IMPORTANT] > Better using New version of ChemLLM! > AI4Chem/ChemLLM-7B-Chat-1.5-DPO or AI4Chem/ChemLLM-7B-Chat-1.5-SFT

Saves the delayed operations of a DelayedArray to a HDF5 file. This enables efficient recovery of the DelayedArray's contents in other languages and analysis frameworks.

Tools for data management, count preprocessing, and differential analysis in massively parallel report assays (MPRA).

Chemma-2B is a continually pretrained gemma-2b model for organic molecules. It is pretrained on 40B tokens covering 110M+ molecules from PubChem as well as their chemical properties (molecular weight, synthetic accessibility score, drug-likeness etc.) and similarities (Tanimoto distance between…

## Quick Start ```Python from transformers import AutoTokenizer, AutoModel

zitools allows for zero inflated count data analysis by either using down-weighting of excess zeros or by replacing an appropriate proportion of excess zeros with NA. Through overloading frequently used statistical functions (such as mean, median, standard deviation), plotting functions (such as boxplots or heatmap) or differential abundance tests, it allows a wide range of downstream analyses for zero-inflated data in a less biased manner. This becomes applicable in the context of microbiome analyses, where the data is often overdispersed and zero-inflated, therefore making data analysis extremly challenging.

A R interface to the TnT javascript library (https://github.com/ tntvis) to provide interactive and flexible visualization of track-based genomic data.

Universal chart comprehension and reasoning model

HybridExpress can be used to perform comparative transcriptomics analysis of hybrids (or allopolyploids) relative to their progenitor species. The package features functions to perform exploratory analyses of sample grouping, identify differentially expressed genes in hybrids relative to their progenitors, classify genes in expression categories (N = 12) and classes (N = 5), and perform functional analyses. We also provide users with graphical functions for the seamless creation of publication-ready figures that are commonly used in the literature.

Utility that performs integrated analyses of 'gene' data (a set of genes or other genomic features) with 'peak' data (a set of regions, for example ChIP peaks) to identify the genes nearest to each peak, and vice versa.

CellMixS provides metrics and functions to evaluate batch effects, data integration and batch effect correction in single cell trancriptome data with single cell resolution. Results can be visualized and summarised on different levels, e.g. on cell, celltype or dataset level.

The academic event ontology, currently still in development and thus unstable, is an OBO compliant reference ontology for describing academic events such as conferences, workshops or seminars and their series. It is being developed as part of the [ConfIDent project](https://projects.tib.eu/confident/) to allow RDF representations of the academic events and series stored and curated in the [ConfIDent platform](https://www.confident-conference.org/index.php/main_page).

A python package for optimizing chemical reactions using machine learning (contains 10 algorithms + several benchmarks).

# Mistral-7B-DSM5 This model is fine-tuned from the Mistral-7b-Instruct-v0.2. We propose a method where an Instruct-Tuned language model first acquires knowledge in a specific domain through RAG (Retrieval-Augmented Generation) techniques.

A collection of microRNAs/targets from external resources, including validated microRNA-target databases (miRecords, miRTarBase and TarBase), predicted microRNA-target databases (DIANA-microT, ElMMo, MicroCosm, miRanda, miRDB, PicTar, PITA and TargetScan) and microRNA-disease/drug databases (miR2Disease, Pharmaco-miR VerSe and PhenomiR).

Transform arXiv research papers into engaging presentations and YouTube-ready videos

Batteries included genomic analysis pipeline for variant and RNA-Seq analysis, structural variant calling, annotation, and prediction.

!alt text

HERON is a software package for analyzing peptide binding array data. In addition to identifying significant binding probes, HERON also provides functions for finding epitopes (string of consecutive peptides within a protein). HERON also calculates significance on the probe, epitope, and protein level by employing meta p-value methods. HERON is designed for obtaining calls on the sample level and calculates fractions of hits for different conditions.

DifferentialRegulation is a method for detecting differentially regulated genes between two groups of samples (e.g., healthy vs. disease, or treated vs. untreated samples), by targeting differences in the balance of spliced and unspliced mRNA abundances, obtained from single-cell RNA-sequencing (scRNA-seq) data. From a mathematical point of view, DifferentialRegulation accounts for the sample-to-sample variability, and embeds multiple samples in a Bayesian hierarchical model. Furthermore, our method also deals with two major sources of mapping uncertainty: i) 'ambiguous' reads, compatible with both spliced and unspliced versions of a gene, and ii) reads mapping to multiple genes. In particular, ambiguous reads are treated separately from spliced and unsplced reads, while reads that are compatible with multiple genes are allocated to the gene of origin. Parameters are inferred via Markov chain Monte Carlo (MCMC) techniques (Metropolis-within-Gibbs).

Materials informatics benchmark

SGC is a semi-supervised pipeline for gene clustering in gene co-expression networks. SGC consists of multiple novel steps that enable the computation of highly enriched modules in an unsupervised manner. But unlike all existing frameworks, it further incorporates a novel step that leverages Gene Ontology information in a semi-supervised clustering method that further improves the quality of the computed modules.

Web application for LLM-assisted manuscript review and annotation

Convert PDF files into editable slides with three lines of code

TinyDNABERT is a specialized deep learning model designed for understanding the language of DNA and performing DNA sequence classification tasks. This model is a compact and efficient version of the DNABERT model, optimized to reduce memory usage while maintaining high performance.

Deep learning-based protein sequence design (inverse folding) from backbone structures, achieving 52.4% sequence recovery vs 32.9% for Rosetta, core tool in modern protein design pipelines (Baker Lab, Science 2022)

Workflow library embedded in the Go programming language, focusing on supporting complex workflow constructs, compiling to a single binary, providing powerful file naming and comprehensive audit reports for every output

Structure-aware prefix adaptation for integrating LLMs with knowledge graphs (ACM MM 2024)

Large-scale table detection and recognition dataset with pre-trained models

Powerful and flexible machine learning platform for drug discovery, providing comprehensive tools for molecular property prediction, generative models, knowledge graph reasoning, and reaction prediction with PyTorch backend (1.5K+ stars)

ChemFIE-SA is a BERT-like sequence classifier for predicting synthesis accessibility given a SELFIES string of a compound, fine-tuned from gbyuvd/chemselfies-base-bertmlm on DeepSA's expanded dataset from Wang et al. 2023.

This model is a BERT-like sequence classifier for 221 human protein drug targets, fine-tuned from gbyuvd/chemselfies-base-bertmlm on a dataset derived ChemBL34 (Zdrazil et al. 2023). It predicts potential drug targets using chemical structures represented as SELFIES (Self-Referencing Embedded…

A concept scheme that defines the types of relationships between a learning resource and a node in an educational framework.

The *MungeSumstats* package is designed to facilitate the standardisation of GWAS summary statistics. It reformats inputted summary statisitics to include SNP, CHR, BP and can look up these values if any are missing. It also pefrorms dozens of QC and filtering steps to ensure high data quality and minimise inter-study differences.

Perform the zFPKM transform on RNA-seq FPKM data. This algorithm is based on the publication by Hart et al., 2013 (Pubmed ID 24215113). Reference recommends using zFPKM > -3 to select expressed genes. Validated with encode open/closed chromosome data. Works well for gene level data using FPKM or TPM. Does not appear to calibrate well for transcript level data.

Comprehensive collection of Chinese medical datasets for AI research

Peptide Set Test (PepSetTest) is a peptide-centric strategy to infer differentially expressed proteins in LC-MS/MS proteomics data. This test detects coordinated changes in the expression of peptides originating from the same protein and compares these changes against the rest of the peptidome. Compared to traditional aggregation-based approaches, the peptide set test demonstrates improved statistical power, yet controlling the Type I error rate correctly in most cases. This test can be valuable for discovering novel biomarkers and prioritizing drug targets, especially when the direct application of statistical analysis to protein data fails to provide substantial insights.

Resources on ChIP-seq data which include papers, methods, links to software, and analysis.

The Mistral-DNA-v1-138M-bacteria Large Language Model (LLM) is a pretrained generative DNA text model with 17.31M parameters x 8 experts = 138.5M parameters. It is derived from Mistral-7B-v0.1 model, which was simplified for DNA: the number of layers and the hidden size were reduced.

Model Card for "medllama" ---------------------------

Despite the recent advances of modern GWAS methods, it still remains an important problem of addressing calculation an effect size and corresponding p-value for the whole gene rather than for single variant. The R- package rqt offers gene-level GWAS meta-analysis. For more information, see: "Gene-set association tests for next-generation sequencing data" by Lee et al (2016), Bioinformatics, 32(17), i611-i619, <doi:10.1093/bioinformatics/btw429>.

UNIX-style FASTA manipulation tools.

This is a ReactionT5 pre-trained to predict the products of reactions.

Quantitative DNA sequencing for chromosomal aberrations. The genome is divided into non-overlapping fixed-sized bins, number of sequence reads in each counted, adjusted with a simultaneous two-dimensional loess correction for sequence mappability and GC content, and filtered to remove spurious regions in the genome. Downstream steps of segmentation and calling are also implemented via packages DNAcopy and CGHcall, respectively.

The Semantic Web for Earth and Environmental Terminology is a mature foundational ontology that contains over 6000 concepts organized in 200 ontologies represented in OWL. Top level concepts include Representation (math, space, science, time, data), Realm (Ocean, Land Surface, Terrestrial Hydroshere, Atmosphere, etc.), Phenomena (macro-scale ecological and physical), Processes (micro-scale physical, biological, chemical, and mathematical), Human Activities (Decision, Commerce, Jurisdiction, Environmental, Research).

Biomedical text generation