We help scientists analyze the most relevant gene expression data for their questions
Our mission is to enable scientists to treat and cure disease
We build software to enable biologists to use their knowledge and expertise to discover and validate targets via the analysis and exploration of highly relevant gene expression datasets. Founded by Brian Fox LinkedIn
Personalized Science: Every scientific team has a different focus. We are not aiming to make every piece of expression data available, but rather the key, high-quality datasets in your area of research. It is up to you how personalize your server instance with specific datasets and analyses.
Our gene expression app is personalized for you.
Designed with the inquisitive biologist in mind. We focus on gene expression (bulk and single cell) because of the high re-use value and it is fundamentally approachable compared to other omics data types.
Use any laptop to browse the data. No need for users to install anything.
Cloud or on-prem
Use a shared cloud instance or a custom install on servers you control, within your firewall. The server app is built using django and django rest framework and can be installed via docker or we can provide a pre-configured EBS volume.
Most of the plots and many of the analyses use R and various open-source and community built packages: no secret methods. Although we use ggplot, we do NOT use R-Shiny for the user interface. The R API allows you to perform your own analyses in R and easily grab any data from the server database.
We have a library of over 100 public datasets and over 1M cells that you can select for your instance of the app. The app has curation tools for adding new GEO datasets very quickly and easily.
Your needs can be prioritized in upcoming development sprints of the app. We love working closely with fellow scientists and solving their problems.
There are numerous ways to analyze and curate the datasets in the app.
Jitter and BoxplotThe main boxplot page gives you a quick understanding of the samples in the dataset. It also provides access to other useful data for this gene in this dataset, including saved scores, correlation search, dataset details, gene details. These are fully configurable - can change groupings, colors, subgroups, paired lines.
Gene detailsThe main point-of-interest on the Gene page is the "expression scores" section which shows the most relevant datasets for this gene. Gene descriptions, aliases and structural features are collated from Entrez Gene and Uniprot.
Best ranking scoresZooming in on the expression score table for each gene: the datasets at the top are those where this gene has the largest differential expression, mean or stdev (or other scores that you want to generate).
Correlation SearchA truly unique feature is the ability to find genes that are correlated to the current gene in the dataset. This is dynamically calculated and only takes a few seconds. Knowing which genes are correlated across the samples helps to paint a fuller picture of the biology going on in the system.
PCAPrincipal Component Analysis is a foundational and quick way to explore sample relationships, identify outliers, and estimate the major sources of variation in a dataset.
View UMAP and tSNE plotsExplore scRNA-seq data by looking at cell cluster labels and how any gene is expressed across the UMAP or tSNE space.
Run limmaRun limma or voom+limma on your gene expression and save the differential scores. These saved scores are for your users to browse.
Search Differential GenesTarget Discovery: search for genes which are differential in multiple datasets and have other structural features like transmembrane domains or signal peptides.
Search ResultsThe search results provide links directly to the queried datasets. The resulting lists can be sent to a heatmap or gene set enrichment.
Composition plotParticulary useful for scRNA-seq datasets, the composition plot shows the proportion of samples (cells in this case) within each cell type which are from each donor. The is helpful for understanding the big picture of the dataset.
Custom plotsYou can provide you own R script for plotting or analyzing a particular dataset in a certain way to your liking. This is a valuable way for the computational biologist who does the main analysis of the dataset to share a specific type of plot for other users (or themselves).
Protein sequence and featuresThe protein is shown using colors based on the amino acid properties in order to quickly spot hydrophboic regions, repetetive regions, etc. Data from uniprot is also shown to highlight structural domains.
- 01. Patients
- We really want you to succeed in making drugs or improving patients lives. This is our top priority
- 02. Discovery
- Target discovery is hard and a rare gift when it succeeds. We want to give more scientists the chance to find something new
- 03. Seattle
- We started a few miles from the iconic Needle in Seattle. We have global reach
The latest news in our blog
Brian Fox • Feb 3 2023
[reserved for future blog posts]
Reach out for a demo or licensing information. We won't spam you. We don't have pushy sales (we want to make sure your use case is a good fit)
[submit form first]
brian at needlegenomics dot com