More about the CellKb platform

Explore our latest blog posts, webinars and case studies to learn how CellKb can advance your research.

Latest Blog Posts

Deep dives into methods and comparisons

Expert Webinars

Live sessions and tutorials on CellKb features

Use Cases

Real-world applications and success stories

More about the CellKb platform

Explore our latest blog posts, webinars and case studies to learn how CellKb can advance your research.

Latest Blog Posts

Deep dives into cell type annotation methods and comparisons

Expert Webinars

Live sessions and tutorials on CellKb features

Use Cases

Real-world applications and success stories

More about the CellKb platform

Explore our latest blog posts, webinars and case studies to learn how CellKb can advance your research.

Blog Posts

Webinars

Use Cases

More about the CellKb platform

Explore our latest blog posts, webinars and case studies to learn how CellKb can advance your research.
How to cite CellKb
Website

Please cite CellKb with a link to https://www.cellkb.com.

Preprint

CellKb Immune: a manually curated database of mammalian hematopoietic marker gene sets for rapid cell type identification. Ajay Patil & Ashwini Patil, bioRxiv 2020.12.01.389890, 2020. doi: 10.1101/2020.12.01.389890

Recent Blog Posts
Explore our latest articles on cell type annotation and analysis
Comparison
Traditional and AI annotation methods

How AI cell type annotation methods compare to traditional approaches and what makes CellKb unique.

21 April 2025 Read article →
Case Study
Granular cell type annotations with CellKb

Predicting granular cell types in the Tabula Muris Senis mouse liver single-cell dataset and comparing with CellTypist.

10 October 2024 Read article →
Case Study
Cell type prediction using CellKb

Annotating individual cells in the PBMC dataset with CellKb and comparing with SingleR.

11 January 2024 Read article →
Tutorial
Cluster annotation with CellKb

Cell type annotation of clusters in PBMC single-cell RNA-seq data and how it compares with SingleR and PanglaoDB.

20 June 2023 Read article →
Webinars & Videos
Watch our in-depth tutorials and demonstrations
Tutorial
scRNA-seq data analysis and cell type annotation using CellKb

Webinar hosted by the Single-cell Immunophenotyping Core at the University of Chicago on the basics of scRNA-seq data analysis followed by cell type annotation with CellKb.

60 minutes Watch now →
Integration
Cell type prediction using CellKb in BioBam OmicsBox

Webinar hosted by BioBam on the integration of CellKb within OmicsBox. Cluster annotation can now be performed in OmicsBox by directly calling the CellKb API.

60 minutes Watch now →
Award
Best of Show award for CellKb

CellKb won the Best of Show Award at Bio-IT World Expo 2024 in Boston. Our CEO talks about what makes CellKb a standout solution for cell type annotation and biomarker discovery.

1 minute Watch now →
Use Cases
See how researchers use CellKb for advanced cell type annotation and analysis
Overview

Cell type prediction for 500,000 spots in custom whole mouse spatial RNA-seq data.

Challenge
  • Creating a reference of cell types from 16 tissue types in mouse.
  • Standardizing tissue and cell type annotations from all references.
  • Limited computational resources available for cell type prediction.
  • Solution
  • CellKb provided a reference dataset of ~19K gene signatures representing ~500 cell types from 16 tissues taken from 450+ publications.
  • The reference data consisted of harmonized cell type and tissue ontologies.
  • Optimized cell type prediction algorithm enabled rapid and accurate identification of cell types from the database, minimizing computational resource requirements.
  • Overview

    Absence of a suitable reference database for annotating cell types in non-model organisms.

    Challenge
  • Integrating multiple datasets with different levels of cell type granularity to create a comprehensive reference is challenging.
  • Limited reference data with broad cell types only, eg. Hepatocytes, Macrophages.
  • Unable to identify sub-types and states of cell types of interest.
  • Solution
  • CellKb uses curated cell type signatures from published studies as references, eliminating the need to integrate datasets and removing batch effects.
  • The CellKb knowledgebase contains a comprehensive collection of broad and granular cell types along with their hierarchical relationships and experimental conditions.
  • Detailed subtypes such as "Periportal hepatocytes" and "Alveolar macrophages", as well as cell states like "activated" and "quiescent" are predicted using CellKb.
  • Overview

    Replacing broad cell type annotations with granular cell types using a comprehensive reference.

    Challenge
  • Unable to reliably annotate cell types in non-model organisms with high confidence.
  • Using multiple references independently results in different annotations for the same cells due to lack of harmonization.
  • Solution
  • CellKb annotates non-model organisms with the help of cell type references from model organisms and its cross-species prediction capabilities.
  • Quality of cell type signatures in CellKb is verified using reliability scoring metrics.
  • Overview

    Confirming cell type predictions made by AI methods.

    Challenge
  • Multiple AI methods using different reference models provide varying levels of cell type predictions.
  • AI models are not regularly updated with the latest publications resulting in inaccurate annotations.
  • AI methods produce black box predictions, lacking transparency about the reference data and criteria used for cell type assignment.
  • Solution
  • CellKb performs cell type prediction based on matching cell type signatures in literature and canonical markers providing a complementary approach to AI prediction methods.
  • The CellKb knowledgebase is regularly updated and includes cell types from the latest publications.
  • CellKb gives the details of each publication to which a cell type prediction is linked, enhancing transparency and trust in the annotations.
  • Try CellKb Today

    Start annotating your single-cell data with confidence