AtlasXomics LatchBio Tutorials

Introduction

AtlasXomics has partnered with LatchBio to provide easy to use workflows to quickly process our spatial epigenomics data. Below is a summary of that process:

Workflow input tutorials

Walkthroughs below take you step by step through running our workflows. For more information on each workflow you can also go to its about page on LatchBio.

Processing Data:

Whether reprocessing data ran through our service or processing data generated in your lab from our kits, these workflows can be used to analyze the data from fastqs to our R Shiny app and other downstream analyses.

If starting from fastq files first run our ATX epigenomic preprocessing workflow:

To quickly check the quality of your spatial data and optimize any filtering or clustering next run our optimize archR workflow. If you are using our kits, make sure you have already ran AtlasXBrowser (see pods section) to get the spatial folder needed for running opitmize archR.

For some samples, it can be helpful to clean the samples of any high count lanes with the clean workflow. While optional, this can help reduce clustering artifacts. In our service data if you received cleaned_fragment files this was already completed. It can be helpful to run this if you are seeing lane artifacts in your optimize archR workflow.

Once you are happy with the clustering, filtering and/or cleaning of your samples, running the create ArchRProject workflow will fully process the data (including peak calling and motif enrichment analysis), creating the files needed necessary for running the shiny app as well as the archRProject, Seurat Objects and bigwigs for further analysis. Many figures and tables are also created to look at the marker genes, peaks and motifs for the indicated clustering parameters.

Downstream analyses:

After the ArchRProject has been created, the shiny app can be setup in a pod and explored. (shiny app instructions in the following section). Additional downstream analyses are also available for some common applications:

For investigating the spatial distributions of clusters use the neighborhood analysis workflow. This workflow based on Bäckdahl 2021 identifies how often each cluster is adjacent to other clusters in a given dataset. Note: this workflow currently can not support greater than 1 FG220 run.

For identifying differences between particular cluster sets use the compare cluster workflow. This workflow takes in 2 sets of clusters as an input and identifies the differential genes, peaks and motifs between the datasets. Make sure to check out the about page of this workflow for the naming rules for the groups.

For large datasets or gene lists the module score workflow takes in a provided .csv file for a list of genes and generates a spatial signature score with that set of genes for every sample in the dataset. Outputs show the scores scaled individually by sample or scaled together for comparison. This typically used to help with cluster annotation by providing multiple genes of a cell type as the module instead of just singular marker genes.

Pod instructions

ATX R Shiny App for Epigenomics

To quickly visualize and explore your data, our R Shiny app is available to get you started on your analyses. Whether you are using a new app or starting your analysis on a pre-generated service app the guidelines below should get you started.

AtlasXBrowser

AtlasXBrowser locates where on a tissue image that our spatial epigenomics assay was performed as well as indicating which datapoints don't have tissue and should be fitlered out. This software creates the spatial folder which is used in many workflows to display the data spatially.

To set up AtlasXbrowser in LatchBio follow the instructions here and for running the software whether locally or on LatchBio go here. For LatchBio users you can skip the installation step.