- 2021-09-07 15:00, Yu-jian Kang
- This is a repository of RiboCalc data and scripts.
Dependencies:
Model building: R packages - caret, glmnet
Feature caculation scripts: Python packages - biopython, numpy
Feature caculation tools: fimo, RNAfold, TargetScan (miRbase), bedtools, MTDRcalculator, EMBOSS
Human model comparison: Python packages - pdb, pandas, numpy, sklearn, scipy, matplotlib, seaborn
tom@linux$ git clone git@github.com:gao-lab/RiboCalc.git
Taking OCTOPOS raw data as an example (Figure S10):
The scripts for feature calculation are at feature_calculation/script
The public datasets are provided in feature_calculation/raw_data
The calculated feature value of OCTOPOS is at feature_calculation/feature_data/
tom@linux$ cd feature_calculation/script
tom@linux$ sh test_OCTOPOS.sh
RiboCalc model building(Figure 2A-C, Figure 3A):
The RiboCalc model data is RiboCalc/RiboCalc.RData
To build RiboCalc without RNA expression, see RiboCalc/remove_RNAexpression.r
The prediction result for TE is TE_testing_result.tab
tom@linux$ cd RiboCalc
tom@linux$ Rscript RiboCalc_build_model.r
Build cell specific models for the 5 cell lines (Table 1, Figure S6)
tom@linux$ cd cell_specific_model
tom@linux$ Rscript cell_specific_model.r
Performance comparison with LiJJ's human model and SamplePJ's model (Table S6, Figure S7)
tom@linux$ cd human_model_comparison/LiJJ
tom@linux$ Rscript testing_LiJJ_human.r
tom@linux$ cd ../SamplePJ/script
tom@linux$ sh testing_SamplePJ.sh
RiboCalc performance testing in yeast (Figure 2D-E, Table 3)
tom@linux$ cd RiboCalc_yeast
tom@linux$ Rscript RiboCalc_yeast_model.r
RiboCalc prediction of lncRNAs binding with ribosomes reported by previous studies (Figure 3B-C)
tom@linux$ cd ribo-lncRNA
tom@linux$ Rscript ribo_lncRNA.r
If you have any questions about RiboCalc, please mail ribocalc@mail.cbi.pku.edu.cn.