search

Zafar-Lab / Margaret

Metric learning-based graph-partitioned trajectory inference from single-cell data
MIT License
11 stars 5 forks source link

provide the ipynb file of multifurcating and placenta Trophoblast differentiation, Mouse cell atlas, Oligodendrocyte differentiation, and Planaria parenchyme differentiation #39

Open YUYING07 opened 9 months ago

YUYING07 commented 9 months ago

Can you provide the ipynb file for placenta Trophoblast differentiation, Mouse cell atlas, Oligodendrocyte differentiation, and Planaria parenchyme differentiation and multifurcating in the simulated dataset?

  1. I get the number of clusters for multifurcating based on the code in metric_sim_disconnected9 and the results are very different from the paper. And it is highly affected by the train_metric parameter.
  2. placenta Trophoblast differentiation, Mouse cell atlas, Oligodendrocyte differentiation, and Planaria parenchyme differentiation have been log-transformed, so is it still necessary to run preprocess_recipe?
kpandey008 commented 9 months ago

Hi YU,

1) Can you share the dataset name and your current results for the multifurcating dataset.

2) For the other 4 real datasets, you dont need to perform any pre-processing. Please refer to the following colab notebooks for different parameters used:

Placenta Trophoblast differentiation: https://colab.research.google.com/drive/17JCRlzUxc1nVyx2qEZ9hA9ewahosoKmQ?usp=sharing

Mouse cell atlas: https://colab.research.google.com/drive/1glQMPN3UWdIZIESi6m68FHDe02ZPlczj?usp=sharing

Oligodendrocyte differentiation: https://colab.research.google.com/drive/18KRbu2HkOCfxflc1PGYkwMjzLmGrx6XU?usp=sharing

Planaria parenchyme differentiation: https://colab.research.google.com/drive/1Jk8vqsxiyz-aVa5jQOTP-EfNd-9CJWbV?usp=sharing

Best, Kushagra