Can user train models using new data sets?

[xiekunwhy]

Hi,

TransposonUltimate (https://academic.oup.com/nar/article/50/11/e64/6541023) authors collected a lot repeat sequences, can users train models using these dataset to build new models using DeepTE? How to train if we can?

Best Kun

[songliVT]

Yes, we do have training script for DeepTE, that can be used to train new models.

Song

On Fri, Aug 12, 2022 at 3:49 AM xiekunwhy @.***> wrote:

Hi,

TransposonUltimate (https://academic.oup.com/nar/article/50/11/e64/6541023) authors collected a lot repeat sequences, can users train models using these dataset to build new models using DeepTE? How to train if we can?

Best Kun

— Reply to this email directly, view it on GitHub https://github.com/LiLabAtVT/DeepTE/issues/20, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACEEENUKTDHH2KUIBUVYP6TVYX6YRANCNFSM56KYQ5XA . You are receiving this because you are subscribed to this thread.Message ID: @.***>

-- Associate Professor in Plant Genomics and Bioinformatics School of Plant and Environmental Sciences Virginia Polytechnic Institute and State University

zoom https://virginiatech.zoom.us/j/8790572835

[xiekunwhy]

Do you mean scripts in this directory (https://github.com/LiLabAtVT/DeepTE/tree/master/example_train_script)? But I really don't know how to use them, would you please add some introductions?

Best, Kun

[yanhaidong1]

Hey Kun,

I updated a training_example_dir that helps users to do the training using their own data. Please download the latest version.

You can run by following the guideline. If you want to train your own data, I suggest you need to modify a little bit for the TE names and classes shown in the script 'classify_TE_keras_model_predict_kmer.py' and 'one_hot_rep_kmer.py' to match your own cases.

Basically, DeepTE does not provide a professional training process in the current version, but if you want to do the training, the scripts attached in the training_example_dir would be really helpful.

Hopefully, we will update a version of DeepTE in the future that will be easier for users to have a training.

Best wishes, Haidong

[xiekunwhy]

Hi Haidong,

Thank you for your reply, I will try to train models following your suggestions.

Best regards, Kun

[davidaray]

I just tried to download the examples but am getting this error. Is it possible the file has been corrupted?

$ tar -xvf ipt_shuffle_All_CNN_data.tar.gz

gzip: stdin: not in gzip format tar: Child returned status 1 tar: Error is not recoverable: exiting now