Deconer: An Evaluation Toolkit for Reference-based Deconvolution Methods Using Gene Expression Data.

Area of Science:

• Computational biology
• Bioinformatics
• Genomics

Background:

• Reference-based deconvolution methods accurately quantify cell type proportions from transcription data.
• A lack of comprehensive evaluation and guidance exists for these methods.

Purpose of the Study:

• Introduce Deconvolution Evaluator (Deconer), a toolkit for evaluating reference-based deconvolution methods.
• Provide systematic comparisons and insights into cell proportion deconvolution algorithms.

Main Methods:

• Developed Deconer with simulated and real gene expression datasets (bulk and single-cell).
• Conducted systematic comparisons of 16 deconvolution methods.
• Analyzed method robustness, rare component deconvolution, signature gene selection, and external reference building.

Main Results:

• Performed in-depth analysis of application scenarios and challenges.
• Provided constructive suggestions for selecting and developing deconvolution algorithms.
• Demonstrated Deconer's utility in comparing diverse deconvolution methods.

Conclusions:

• Deconer offers valuable insights for researchers choosing deconvolution tools.
• Facilitates clinical applications and advances deconvolution tool development for gene expression data.
• The toolkit, code, and data are publicly available.