Key takeaways

  • iSCanGuide™ v3.0 scales to tens of millions of cells and adds atlas-scale AI-model-based annotation and a novel spatially aware trajectory inference method.
  • In the live webinar demo, AdvaitaBio bioinformatics scientist Beverly Robinson analyzed a 465,534-cell clear cell renal cell carcinoma sample profiled on the Xenium in situ platform.
  • The complete analysis, from raw data upload to trajectory analysis, took about 20 minutes of compute time.
  • In the tumor epithelium, differential expression identified 140 significant genes, with EGFR as a key marker, and epithelial cells were the hub of a signaling network with 819 interactions.
  • You can book a live demo to see iSCanGuide™ v3.0 run on your own spatial or single-nucleus data.

Our iSCanGuide™ v3.0 webinar packed a lot into 45 minutes, centered on a live iSCanGuide™ demo. Beverly Robinson, one of our bioinformatics scientists here at AdvaitaBio, conducted a live single-cell spatial analysis from raw data upload to trajectory analysis. This recap is for anyone who missed this single-cell analysis and spatial transcriptomics webinar, anyone comparing platforms for spatial transcriptomics, and anyone giving iSCanGuide™ a second look.

iSCanGuide™ v3.0 is redesigned to make sophisticated single-cell and spatial analysis faster and more accessible. That applies whether you’re an experienced computational researcher or a scientist who doesn’t have the time, resources, or specialized expertise to build and manage these workflows yourself.

What’s new in iSCanGuide™ v3.0?

iSCanGuide™ v3.0 is a redesign of AdvaitaBio’s single-cell and spatial transcriptomics analysis platform. The v3.0 release adds scale to tens of millions of cells, a simpler experience for less-experienced users, faster iteration for experienced users, analysis provenance and sharing, spatially aware trajectory inference, and atlas-scale AI-model-based annotation.

Scale to tens of millions of cells. v3.0 scales to tens of millions of cells, with significantly improved UI performance, so large datasets stay responsive as you explore them.

A redesigned experience for less-experienced users. The new design lets less-experienced users run meaningful projects without ongoing supervision or support.

Faster, more iterative analysis for experienced users. If you could already run these analyses yourself, v3.0 cuts the time they take, so you can test more clustering resolutions, annotations, and hypotheses in the same session.

Advanced analysis provenance and sharing. v3.0 adds more advanced provenance tracking and sharing, so you can see how a result was produced and share it with collaborators.

A novel spatially-aware trajectory inference method. This new method brings tissue context into trajectory analysis.

AI-model-based annotation at atlas scale. v3.0 brings AI-model-based cell annotation to atlas-scale datasets.

What did the iSCanGuide™ v3.0 webinar demo show?

In the live webinar demo, Beverly Robinson analyzed a publicly available clear cell renal cell carcinoma sample with 465,534 segmented cells, profiled on the Xenium in situ platform. The complete analysis took about 20 minutes of compute time.

The demo workflow at a glance:

  1. Upload the raw Xenium data; iSCanGuide™ automatically identifies the metadata, expression matrices, and H&E images.
  2. Apply QC filters for transcripts per cell and nuclei per cell.
  3. Build a baseline embedding and run Leiden clustering at resolution 0.6.
  4. Annotate cell types with SingleR and view them on the tissue image.
  5. Run differential expression and cell-cell communication analysis on the tumor epithelium.
  6. Separate healthy from tumor cells with gene-signature enrichment, then run pseudotime trajectory analysis.

Setup with minimal manual work. After upload, iSCanGuide™ automatically identified the metadata, expression matrices, and H&E images. Beverly adjusted QC filters for transcripts per cell and nuclei per cell.

Baseline embedding and annotation. She built a baseline embedding from the full gene panel and explained why sample integration matters for multi-sample projects and how to distinguish donor effects from true biological cell states. She chose Leiden clustering at resolution 0.6 (Louvain and K-means are also available) and annotated cell types with SingleR and a reference dataset. Viewing the annotations on the tissue image revealed distinct biological regions and cell populations.

Tumor epithelium biology. Focusing on about 30,000 epithelial cells at the tumor’s center:

  • Wilcoxon differential expression identified 140 significant genes, with EGFR as a key marker.
  • Cell-cell communication mapping showed epithelial cells as the hub of the tissue’s signaling network, with 819 interactions, including HLA-DRB1 to CD4.
  • A second embedding on epithelial cells alone resolved subtypes within that population.

Healthy versus tumor cells and trajectory analysis. Beverly uploaded gene signatures for healthy kidney tubular epithelium and terminal states, and enrichment annotation separated healthy tubular cells (cluster 2) from terminal, proliferative states (clusters 4 and 5). A pseudotime trajectory then traced progression from the most proliferative to the least transformed cells.

Frequently asked questions

What did the iSCanGuide™ v3.0 webinar cover?

The iSCanGuide™ v3.0 webinar covered what’s new in v3.0 and featured a live demo by AdvaitaBio bioinformatics scientist Beverly Robinson. In the demo, Beverly ran a complete single-cell spatial analysis, from raw data upload through QC, clustering, cell type annotation, differential expression, cell-cell communication, and trajectory analysis.

What dataset was analyzed in the iSCanGuide™ v3.0 demo?

The demo used a publicly available clear cell renal cell carcinoma (ccRCC) sample with 465,534 segmented cells, profiled on the Xenium in situ platform. The complete analysis in iSCanGuide™ v3.0 took about 20 minutes of compute time.

Is the iSCanGuide™ v3.0 webinar recording available?

The iSCanGuide™ v3.0 webinar recording is available on YouTube here.

How can I see iSCanGuide™ v3.0 with my own data?

You can book a live demo with the AdvaitaBio team. We’ll walk through your own spatial or single-nucleus data in iSCanGuide™ v3.0.

Is iSCanGuide™ v3.0 right for your evaluation?

If you’re comparing spatial analysis tools, iSCanGuide™ v3.0 gives you a few concrete things to test: how it handles your largest dataset, how quickly a newer team member can run a project unsupervised, how fast you can iterate, and how easily you can trace and share results.

That was one dataset. We’d love to show you what it looks like with yours. Book a live demo and we’ll walk through your spatial or single nucleus data with you.

Want the full walkthrough first? Watch our webinar recording on YouTube.

References

  1. Aran D, Looney AP, Liu L, et al. Reference-based analysis of lung single-cell sequencing reveals a transitional profibrotic macrophage. Nat Immunol. 2019;20(2):163–172. https://doi.org/10.1038/s41590-018-0276-y
  2. Traag VA, Waltman L, van Eck NJ. From Louvain to Leiden: guaranteeing well-connected communities. Sci Rep. 2019;9:5233. https://doi.org/10.1038/s41598-019-41695-z

Let’s Get Started