ZauggGroup / DeePiCt

Pipeline for the automatic detection and segmentation of particles and cellular structures in 3D Cryo-ET data, based on deep learning (convolutional neural networks).
Apache License 2.0
30 stars 9 forks source link

how to download the demo data in 3D notebook #14

Closed JunMa11 closed 1 year ago

JunMa11 commented 1 year ago

Thanks for sharing the awesome work.

I run into the following error for the 3D jupyter-notebook. How can I download the 180426_005 sample data? It's empty currently.

image

frosinastojanovska commented 1 year ago

Hi @JunMa11, the sample tomogram is provided here: link. You should be able to download the example and run it from the Colab. According to the error, the tomogram should be named as match_spectrum_filt.mrc in your tomo_data Drive folder. So download the tomogram (cropped version for faster execution) and upload it on Drive in the folder tomo_data, also renaming it to match_spectrum_filt.mrc.

JunMa11 commented 1 year ago

Hi @frosinastojanovska ,

Thanks for your guidance very much.

I have another quick question. Which image set is the annotation in the following public data?

https://www.ebi.ac.uk/empiar/EMPIAR-10989/

frosinastojanovska commented 1 year ago

Hi @JunMa11, as I understood from the question, do you want the ground truth annotation for the provided tomograms from the EMPIAR entry? If that is what you look for, you can find it here: image

JunMa11 commented 1 year ago

Hi @frosinastojanovska ,

Thanks for your quick reply very much. It helps me a lot.

Based on the readme file, the following six files are image-label pairs, which can be used to train the segmentation model, right?

image

frosinastojanovska commented 1 year ago

Hi @JunMa11, yes for actin segmentation training those are the pairs of files with data and ground truth labels.

JunMa11 commented 1 year ago

Thank you so much:)