the final results, such as UMAP figure, are not similar to the figures in published paper when I process the data derived from the published paper. what is reason? What is the difference between trailmaker and R, and why do they not get the same results from the same data?
thank you
Hi.
For some context on how Trailmaker works, the platform backend uses R and the Seurat package.
In order to reproduce a published paper, we recommend matching the settings that are reported in the paper, such as for filtering thresholds, integration, UMAP/t-SNE plotting and clustering. If you closely match these settings, you should be able to accurately recreate the results.
Note that the UMAP/t-SNE plot will never be identical due to the inherent random nature of these types of plots.
I hope this answer is helpful. Happy analysing!