Mosaic - python package¶
This package provides a set of tools to analyze data produced by Tapestri instruments (.h5 files). It is NOT open source, although data can be exctrated and analysis can be held either using mosaic functions or custom functions.
package version
This wiki is based on Mosaic v2.2, the package is mantained and constantly updated by missionbio, therefore some things maybe be changed. If you are using a new version of the package please refer to the official documentation.
Installation¶
The package is available in conda.
conda create --name mosaic -c missionbio -c plotly -c conda-forge missionbio.mosaic notebook
conda activate mosaic
Tip - Installing from exported environment
Another way of installing mosaic with some other packages used during the first analysis of the data is with the exported env: /workspace/projects/scell_tall/mosaic180123.txt.
Jupyter tutorials¶
How do I access a jupyter notebook?
please, if you are having hard time accessing a Jupyter notebook, refer to this guide.
Data accessibility
You might need additional accessibility to run some of the analysis available in the jupyter notebooks.
In the folder /workspace/projects/scell_tall/LOPEBIG_44_analysis/ there are four notebooks that were used to do some of the initial analysis on data. In these notebooks you can find some of mosaic built-in functions and some custom functions.
Additionally you can find some other notebooks in the section below: Link and in the folder /workspace/projects/scell_tall/LOPEBIG_44_analysis/
FAQ¶
Since there is no section with FAQ on mosaic documentation, here we collect our questions and the answers missionbio support provided. If you intend to ask more questions to their support, please update this section.
How min_prc_cells and min_mut_prct_cells are computed?
In mosaic they are calculated on the total numbers of cells in the dataset. Differently from Tapestri Insights where they are computed on the percentage of genotyped.
Why do I get different number of variants when uploading samples in Tapestri Insight and when loading a merged .h5 file in Mosaic?
Still waiting for a response
How is the data filtered by the ms.load("path/to/.h5", raw=False, apply_filter=True) function?
- The
rawparameter set to False discards empty barcodes - The
apply_filterparameter set to True loads only the variants that meets the Tapestri Insights advanced filters.
We have a final matrix that excludes all those cells or variants that do not meet the filters and a filter layer called FILTER_MASK filled with 1 and 0 that excludes the genotypes (single cell in the matrix) that do not meet the filters.
Additional resources¶
MissionBio provides some video tutorials on their website: Mosaic tutorials
Additionally, the company provided a personal training course, the video lessons can be found here:
/workspace/projects/scell_tall/LOPEBIG_44_analysis/mosaic/Video_Trainings/MissionBio-3_1-Mosaic.mp4
/workspace/projects/scell_tall/LOPEBIG_44_analysis/mosaic/Video_Trainings/MissionBio-3_2-Mosaic.mp4
Links¶
Reference¶
- Federica Brando
- Raquel Blanco