Overview

Nadiatools performs pre-processing and data processing steps in a scRNA-seq analysis workflow. It is specially designed to analyze data from Nadia instrument, which is based on Drop-seq technology.

What we can do

  • Quality control on sequencing reads

  • Trim adapter sequences and low quality bases

  • Perform read alignment and generate gene expression matrix by STARsolo or Alevin-fry

  • Support single-cell and single-nucleus workflow

  • Perform data processing steps: (1) Calculating quality controls metrics, (2) Doublet detection, (3) Normalizing data, (4) Feature selection, (5) Dimentionality reduction, (6) Visualization (UMAP, T-SNE), (7) Integrating datasets.

Noticeable features

  • Designed for Nadia instrument:
    • Barcode and UMI structures are built-in

    • No barcode whitelist is required

  • Produce an intensive report in html format

  • Produce many useful plots (knee plot, violin plot, PCA, UMAP, T-SNE plots,…)

  • Output files could be used for further downstream analyses (e.g. AnnData object for cellxgene visualization tool, mtx matrix for ASAP tool,…)

Command description

  • nadia-ref: prepare index for alignment. It can generate both STAR index and splici-index for Salmon-Alevinfry workflow.

  • nadia-reads: concatenate reads files across lanes, perform read quality control, trim adapter sequences and low quality bases.

  • nadia-quant: takes reads files from nadia-reads and index from nadia-ref and performs alignment, gene expression quantification and barcode filtering. It supports both STARsolo and Alevin-fry pipeline, single-cell and single-nucleus workflow. nadia-quant produce a feature-barcode matrix and some plots for further processing.

  • nadia-process: takes a matrix (mtx or h5ad) from nadia-quant and performs data processing steps: calculate qc metrics, filter cells, filter genes, doublet detection, normalizing, feature selection, dimentionality reduction and visualization.

  • nadia-combine: takes processed h5ad files from nadia-process and integrates multiple sample together.