Sign inSign up

tblab/nrc

By tblab

•Updated over 9 years ago

nRC: non-coding RNA Classifier

Image
1

1.9K

tblab/nrc repository overview

⁠nRC: non-coding RNA Classifier
⁠This image, based on Centos (7.2.1511) image, contains:
  • Python (2.7.5)
  • Theano (0.8.2)
  • NumPy (1.11.2)
  • Pyyaml (3.12)
  • Java JDK (1.8.0)
  • ViennaRNA (1.8.5)
  • Ipknot (0.0.2)
  • GLPK (4.60)
  • MoSS (2.13)
  • NRC-tool (0.0.1)
⁠nRC-tool source code:

https://github.com/IcarPA-TBlab/nrc⁠

⁠The following command starts a container with the NRC tool.
docker run -d -it --name=<my_container_name> -v <host_experiments_path>/data:/root/nrc_workspace/data tblab/nrc
.
⁠nRC Training process
⁠Step 1: training feature model
docker exec -it <my_container_name> /root/nrc_workspace/nrc_training_feature_model.sh -d <nRNA_training_file>.fasta -o <experiment_name> -n <graph_feature_max_size> -m <graph_feature_min_size>
⁠Step 2: traning CNN (create classification model)
docker exec -it <my_container_name> /root/nrc_workspace/nrc_training_network_model.sh -d <experiment_name>_<graph_feature_max_size>_<graph_feature_min_size>.txt -p <parameters>
⁠.
⁠nRC Testing process
⁠Step 3: testing feature model
docker exec -it <my_container_name> /root/nrc_workspace/nrc_testing_feature_model.sh -d <nRNA_testing_file>.fasta  -f <experiment_name>_<graph_feature_max_size>_<graph_feature_min_size>.nel -o <sequence_output_name>
⁠Step 4: testing CNN (test classification model)
docker exec -it <my_container_name> /root/nrc_workspace/nrc_testing_network_model.sh -d <sequence_output_name> -p <parameters> -m <experiment_name>_<graph_feature_max_size>_<graph_feature_min_size>.pkl -o <classification_output_name>
⁠.
⁠Fasta file format

Note that each sequence in fasta format should have the following header:

>seq_ID class_name

wher seq_ID and class_name labels must be without ".","","/" or any special character, e.g.:

>RF00005_AAFR03000905_1_148681-148750 tRNA
AGCAGTGTGGCATAGTGGAAAGTGTTGGATTTGTAGTTAAAGGACTTGGGTTCAGATCCC
TGCTCTGTTA
⁠.
⁠Example:
  1. Download example data from this link⁠.
  2. Put example data in a folder on your PC, e.g. /home//nrc/data/.
  3. Open a terminal, then execute the following code.
docker run -d -it --name=nrc_tool -v /home/<user>/nrc/data:/root/nrc_workspace/data tblab/nrc

docker exec -it nrc_tool /root/nrc_workspace/nrc_training_feature_model.sh -d sample_train_40.fasta -o experiment1 -n 5 -m 3 

docker exec -it nrc_tool /root/nrc_workspace/nrc_training_network_model.sh -d experiment1_5_3.txt -p parameters.txt

docker exec -it nrc_tool /root/nrc_workspace/nrc_testing_feature_model.sh -d sample_test_20.fasta -f experiment1_5_3.nel -o my_test_seqs

docker exec -it nrc_tool /root/nrc_workspace/nrc_testing_network_model.sh -d my_test_seqs -p parameters.txt -m experiment1_5_3.pkl -o my_first_results.txt
⁠.
⁠Additional dataset:

Dataset used for experiments in the manuscript submitted for pubblication at international journal is available here⁠. It is composed by a training dataset with 6320 ncRNA fasta sequences (belonging to 13 ncRNA classes) and two validation datasets with respectively 2600 fasta sequences (13 classes) and 2400 ncRNA fasta sequences (12 classes). All of them are extracted from Rfam database release 12. This link also contains trained models used in the manuscript submitted for pubblication at international journal.

Tag summary

Content type

Image

Digest

Size

549.8 MB

Last updated

over 9 years ago

docker pull tblab/nrc