SusztakLab Biobank

SusztakLab Biobank · Susztak Laboratory · University of Pennsylvania

Let us cure kidney disease together.

Eight hundred and fifty million people live with kidney disease, and almost none of the treatments we have were designed from human kidney biology. For almost two decades our laboratory has collected and profiled human kidney tissue, paired it with mouse and rat models and with the genetics of millions of people, and built one of the most complete maps of the human kidney that exists. This site is where we share it: genome-wide association results, expression, methylation, protein and open-chromatin maps, single-cell and single-nucleus atlases, spatial transcriptomes and animal-model resources, each in an interactive browser you can query today.

Take the data. Find new biology. Tell us what you discover.

What we share, and why

More than 5,000 human kidneys, read layer by layer.

The SusztakLab Biobank rests on a simple conviction: to understand kidney disease we have to study the human kidney itself. Over two decades we have collected more than 5,000 human kidney samples with clinical information and profiled them at every layer, from genotype and cytosine methylation to open chromatin, gene expression and protein, in bulk tissue, in single nuclei and in intact tissue sections.

The layers answer different questions. Genetics tells us which variants change the risk of disease. The molecular layers tell us in which cell type and through which gene those variants act. Single-cell and spatial maps show how cells change and where they meet in the diseased organ. Mouse and rat models let us test what the human data predict. Each resource on this page is a browser onto one of these layers.

  • Human genetics and QTLsGWAS in 2.2 million people; expression, methylation and protein QTLs in human kidney and blood; the Genetic Scorecard
  • Open chromatin and methylationSingle-nucleus ATAC-seq of healthy and diseased kidneys and cytosine methylation maps, with IGV tracks for every cell type
  • Single-cell atlases12 million cells and nuclei profiled across human, mouse and rat, merged with KPMP and integrated in SISKA
  • Spatial transcriptomicsAdult, diabetic and developing human kidney sections at single-cell resolution
  • Mouse and rat modelsInjury, fibrosis, development, hypertension, diabetes and drug response at single-cell resolution
  • Raw data files81 GEO series, 32 figshare files, 10 Zenodo records and the supplementary tables of every paper, on the Raw data page
30interactive resources
2.2 Mpeople in the kidney function GWAS
5,000+human kidney samples
12 Msingle cells and nuclei profiled
3species: human, mouse, rat
295publications on PubMed

Explore a gene, variant or CpG

Start with what you are studying.

Type a human gene symbol (UMOD), a variant (rs77924615) or a CpG (cg15971010) and we open every browser on this site that can answer for it.

Open for research
Free to use for non-commercial research. Use of any resource on this site is governed by the data use agreement. Please cite the source publication listed on each card and the SusztakLab Biobank (susztaklab.com) in anything you present or publish, and contact Dr. Susztak before publishing results derived from the Data.
Read the agreement

Human genetics and QTL atlases 8

Kidney disease runs in families, and much of that inheritance is carried by common variants scattered across the genome. A genome-wide association study (GWAS) compares millions of these variants between people with better and worse kidney function; in 2.2 million people we found 1,026 regions of the genome where a variant changes kidney function. The catch is that almost none of these variants change a protein: they sit in regulatory DNA, and a GWAS alone cannot say which gene they act on or in which cell. That is what the quantitative trait locus (QTL) atlases answer. In hundreds of human kidneys with both genotype and molecular data we ask, variant by variant, whether a genotype changes the expression of a gene (eQTL, 686 kidneys), the methylation of a CpG site (meQTL, 443 kidneys) or the abundance of a protein (pQTL). When a GWAS signal and a QTL signal point to the same variant, a statistical test called colocalization, the locus gains a gene, a cell type and a mechanism; this is how we built the Kidney Disease Genetic Scorecard, and it is the fastest route we know from a risk variant to a drug target.

Human single-cell and epigenome atlases 7

The kidney is built from more than thirty cell types, and disease does not affect them equally. Single-cell and single-nucleus RNA sequencing measure the genes expressed in each cell separately, so we can see which cell types are lost, which change state (the injured proximal tubule, for example) and which appear in disease, such as the fibrotic microenvironment of immune cells and activated fibroblasts that drives progression. These atlases hold hundreds of thousands of cells from healthy and diseased human kidneys, merged with the KPMP reference and integrated with mouse and rat in SISKA; type a gene and you see in which cell it is expressed. The single-nucleus ATAC-seq atlases add the regulatory layer: they map the open, active regions of the genome in each cell type, up to 237,000 nuclei. Because most disease variants lie in regulatory DNA, these maps tell us in which cell a GWAS variant does its work, and the IGV tracks let you look at any region of the genome, cell type by cell type.

Spatial transcriptomics 3

Single-cell methods dissociate the tissue and lose one essential piece of information: where each cell was. Spatial transcriptomics measures gene expression directly in an intact tissue section, so we can see which cells are neighbours, how a glomerulus, a tubule and the surrounding stroma talk to one another, and how these neighbourhoods reorganise in disease. The viewer holds 22 sections of adult, diabetic and developing human kidney at single-cell resolution. Location turned out to matter: the diabetic atlas revealed a subgroup of diabetic kidney disease defined by B cell-rich immune niches, and the fetal atlas shows how the microenvironment guides progenitor cells to their fate.

Mouse models 8

Human tissue tells us what is associated with disease; animal models let us test cause and effect. In mice we can induce acute injury, fibrosis or diabetes, delete or activate a gene in one cell type, and follow the kidney over time with the same single-cell tools we use in patients. These atlases span the original 2018 single-cell map of the adult mouse kidney, kidney development and its regulatory landscape, and the injury and fibrosis models in which we study how the injured tubule chooses between successful repair and scarring. Comparing them with the human atlases shows which human findings hold in an experimental model, and where mouse and human differ.

Rat models 4

Rats develop hypertension and diabetic kidney disease that resemble the human conditions more closely than most mouse models, and they are large enough for the drug studies that precede a clinical trial. These single-nucleus atlases follow hypertensive and fibrotic (DOCA-salt) and diabetic (ZSF1) rat kidneys, and record what happens in every cell type when the animals are treated with the drugs used in patients: renin-angiotensin-aldosterone blockade, mineralocorticoid receptor antagonists and soluble guanylate cyclase activators. They show not only whether a treatment protects the kidney but how, and in which cells.

How it works

From a question to a discovery in three steps.

We built these resources so that every laboratory, clinic and company can start from the same map of the human kidney that we use.

1

Find your resource

Search by gene, assay or species. Each card opens an interactive browser: gene-level plots, IGV tracks, UMAP viewers or the Samui spatial viewer.

2

Find new biology

Look up your gene, variant or CpG, compare cell types and disease states, test your hypothesis against hundreds of human kidneys before you run a single experiment, and download the tables to go further.

3

Build on it, and cite it

Everything is free for non-commercial research. Cite the paper on each card and susztaklab.com, respect the data use agreement, and write to us when the data lead somewhere new: we love to collaborate.

About

Who we are

The Susztak Laboratory is a kidney genetics and genomics group at the Perelman School of Medicine of the University of Pennsylvania, led by Katalin Susztak, MD, PhD, Willard and Rhoda Ware Professor of Diabetes and Metabolic Diseases and Professor of Medicine and Genetics. The laboratory built one of the largest collections of human kidney tissue assembled for research and has profiled it with bulk, single-cell, single-nucleus, epigenomic, proteomic and spatial methods, pairing the results with large-scale human genetics to connect variants to genes, cell types and mechanisms of chronic kidney disease.

The resources on this site are the interactive companions to those studies. They are maintained by the laboratory within the Renal, Electrolyte and Hypertension Division and the Penn/CHOP Kidney Innovation Center, and they grow as new work is published.

Susztak Laboratory ↗Penn/CHOP Kidney Innovation Center ↗TRIDENT ↗

Members of the Susztak Laboratory
The Susztak Laboratory, Philadelphia