diff --git a/.Rhistory b/.Rhistory
new file mode 100644
index 0000000000000000000000000000000000000000..e69de29bb2d1d6434b8b29ae775ad8c2e48c5391
diff --git a/README.md b/README.md
index 82c136e3ec44eeed2e62f0d1db6fa22e8e2f1ed1..e082aff821db6ebdd396a5ee831c045424943674 100644
--- a/README.md
+++ b/README.md
@@ -1,11 +1,11 @@
written by Niels Schlusser, 15.11.2023
-This repository contains different python scripts to predict translation initiation efficiency from transcript sequences using TranslateLLM, an artificial neural network architecture as presented in "Predicting the translation output from the mRNA sequence - an assessment of the accuracy and parameter-efficiency of deep learning models".
+This repository contains different python scripts to predict translation initiation efficiency from transcript sequences using TranslateLSTM, an artificial neural network architecture as presented in "Predicting the translation output from the mRNA sequence - an assessment of the accuracy and parameter-efficiency of deep learning models".
There are deep learning scripts for essentially three different usecases:
-(1) training a model on synthetic MPRA data in the directory translateLLM_MPRA/
-(2) training a model on endogenous TE data in the directory translateLLM_endogenous/
-(3) do transfer learning from (1) to (2) in the directory tl_TranslateLLM_endogenous/
+(1) training a model on synthetic MPRA data in the directory translateLSTM_MPRA/
+(2) training a model on endogenous TE data in the directory translateLSTM_endogenous/
+(3) do transfer learning from (1) to (2) in the directory tl_TranslateLSTM_endogenous/
Example training data (MPRA from Sample et.al. (2019), endogenous data based on Alexaki et.al. (2020), and clinvar variations based on Landrum et. al. (2020)) are provided in the directory HEK293_training_data/.
Scripts to turn the output of RNAseq and ribosome profiling data into an endogenous data set, appending non-sequential features to a given data set, and constructing a data set based on a vcf file can be found in the directory training_data_preprocessing/.
@@ -19,4 +19,4 @@ The transfer-learning directory also contains a script for making end-to-end pre
For the end-to-end prediction, just the input sequences (UTR and CDS), and, if necessary, the number of exons per transcript need to be provided in a tsv file, all other non-sequential features are computed by the script.
-This code is published under the MIT license.
\ No newline at end of file
+This code is published under the MIT license.
diff --git a/tl_TranslateLLM_endogenous/end2end_prediction.py b/tl_TranslateLSTM_endogenous/end2end_prediction.py
similarity index 100%
rename from tl_TranslateLLM_endogenous/end2end_prediction.py
rename to tl_TranslateLSTM_endogenous/end2end_prediction.py
diff --git a/tl_TranslateLLM_endogenous/predict_clinvar.py b/tl_TranslateLSTM_endogenous/predict_clinvar.py
similarity index 100%
rename from tl_TranslateLLM_endogenous/predict_clinvar.py
rename to tl_TranslateLSTM_endogenous/predict_clinvar.py
diff --git a/tl_TranslateLLM_endogenous/tl_TranslateLLM_HEK293.h5 b/tl_TranslateLSTM_endogenous/tl_TranslateLSTM_HEK293.h5
similarity index 100%
rename from tl_TranslateLLM_endogenous/tl_TranslateLLM_HEK293.h5
rename to tl_TranslateLSTM_endogenous/tl_TranslateLSTM_HEK293.h5
diff --git a/tl_TranslateLLM_endogenous/tl_TranslateLLM_end_HEK293.py b/tl_TranslateLSTM_endogenous/tl_TranslateLSTM_end_HEK293.py
similarity index 96%
rename from tl_TranslateLLM_endogenous/tl_TranslateLLM_end_HEK293.py
rename to tl_TranslateLSTM_endogenous/tl_TranslateLSTM_end_HEK293.py
index 95cd02d0898766b7891cc701ff755acff924746e..80c6ff81a98d51bf58a5335b6b9dad6711e002b5 100644
--- a/tl_TranslateLLM_endogenous/tl_TranslateLLM_end_HEK293.py
+++ b/tl_TranslateLSTM_endogenous/tl_TranslateLSTM_end_HEK293.py
@@ -130,8 +130,8 @@ output_col='TE'
data_path = '../HEK293_training_data/init_effs_HEK293_endogenous.tsv'
scaler_dir = '../HEK293_training_data/scalers/'
-pt_model_path = '../translateLLM_MPRA/TranslateLLM_opt100.h5'
-tl_model_path = 'tl_TranslateLLM_HEK293.h5'
+pt_model_path = '../translateLSTM_MPRA/TranslateLSTM_opt100.h5'
+tl_model_path = 'tl_TranslateLSTM_HEK293.h5'
#nucleotide dictionary
@@ -218,8 +218,8 @@ if len(sys.argv) < 2 or sys.argv[1] == 'predict':
xmin, xmax, ymin, ymax = plt.axis()
plt.text(xmin+1.0, ymax-0.3, '$R_{Pearson}$=%.3f, $R_{Spearman}$=%.3f' % (rho_p,rho_s),fontsize = 12,color='black')
- plt.savefig('scatterplot_tl_TranslateLLM_HEK293.pdf')
+ plt.savefig('scatterplot_tl_TranslateLSTM_HEK293.pdf')
plt.close()
raw_test['predicted_'+output_col] = pred
- raw_test.to_csv("predictions_test_tl_TranslateLLM_HEK293_"+output_col+".tsv",sep="\t",index=False)
+ raw_test.to_csv("predictions_test_tl_TranslateLSTM_HEK293_"+output_col+".tsv",sep="\t",index=False)
diff --git a/translateLLM_MPRA/TranslateLLM_opt100.h5 b/translateLSTM_MPRA/TranslateLSTM_opt100.h5
similarity index 100%
rename from translateLLM_MPRA/TranslateLLM_opt100.h5
rename to translateLSTM_MPRA/TranslateLSTM_opt100.h5
diff --git a/translateLLM_MPRA/TranslateLLM_opt100.py b/translateLSTM_MPRA/TranslateLSTM_opt100.py
similarity index 96%
rename from translateLLM_MPRA/TranslateLLM_opt100.py
rename to translateLSTM_MPRA/TranslateLSTM_opt100.py
index 572b1044e25506db359a4e2687c36ed9c31c558c..99d288221d528278adc60f98f9f10ac49f1f89f3 100644
--- a/translateLLM_MPRA/TranslateLLM_opt100.py
+++ b/translateLSTM_MPRA/TranslateLSTM_opt100.py
@@ -112,7 +112,7 @@ output_col='rl'
data_path = '../HEK293_training_data/opt100_nonseq_feat.tsv'
scaler_dir = '../HEK293_training_data/scalers/'
-integrated_model_path = 'TranslateLLM_opt100.h5'
+integrated_model_path = 'TranslateLSTM_opt100.h5'
#nucleotide dictionary
@@ -175,8 +175,8 @@ if len(sys.argv) < 2 or sys.argv[1] == 'predict':
xmin, xmax, ymin, ymax = plt.axis()
plt.text(xmin+1.0, ymax-0.3, '$R_{Pearson}$=%.3f, $R_{Spearman}$=%.3f' % (rho_p,rho_s),fontsize = 12,color='black')
- plt.savefig('scatterplot_TranslateLLM_opt100.pdf')
+ plt.savefig('scatterplot_TranslateLSTM_opt100.pdf')
plt.close()
raw_test['predicted_'+output_col] = pred
- raw_test.to_csv("predictions_test_TranslateLLM_opt100_"+output_col+".tsv",sep="\t",index=False)
+ raw_test.to_csv("predictions_test_TranslateLSTM_opt100_"+output_col+".tsv",sep="\t",index=False)
diff --git a/translateLLM_endogenous/TranslateLLM_end_HEK293.h5 b/translateLSTM_endogenous/TranslateLSTM_end_HEK293.h5
similarity index 100%
rename from translateLLM_endogenous/TranslateLLM_end_HEK293.h5
rename to translateLSTM_endogenous/TranslateLSTM_end_HEK293.h5
diff --git a/translateLLM_endogenous/TranslateLLM_end_HEK293.py b/translateLSTM_endogenous/TranslateLSTM_end_HEK293.py
similarity index 97%
rename from translateLLM_endogenous/TranslateLLM_end_HEK293.py
rename to translateLSTM_endogenous/TranslateLSTM_end_HEK293.py
index c2d8a43df70f8c74c7f49460deb053a8c9e97fda..ad4d1442b09b56e5d3c41070d5317c6d8200fa41 100644
--- a/translateLLM_endogenous/TranslateLLM_end_HEK293.py
+++ b/translateLSTM_endogenous/TranslateLSTM_end_HEK293.py
@@ -128,7 +128,7 @@ output_col='TE'
data_path = '../HEK293_training_data/init_effs_HEK293_endogenous.tsv'
scaler_dir = '../HEK293_training_data/scalers/'
-integrated_model_path = 'TranslateLLM_end_HEK293.h5'
+integrated_model_path = 'TranslateLSTM_end_HEK293.h5'
#nucleotide dictionary
@@ -197,8 +197,8 @@ if len(sys.argv) < 2 or sys.argv[1] == 'predict':
xmin, xmax, ymin, ymax = plt.axis()
plt.text(xmin+1.0, ymax-0.3, '$R_{Pearson}$=%.3f, $R_{Spearman}$=%.3f' % (rho_p,rho_s),fontsize = 12,color='black')
- plt.savefig('scatterplot_TranslateLLM_end_HEK293.pdf')
+ plt.savefig('scatterplot_TranslateLSTM_end_HEK293.pdf')
plt.close()
raw_test['predicted_'+output_col] = pred
- raw_test.to_csv("predictions_test_TranslateLLM_end_HEK293_"+output_col+".tsv",sep="\t",index=False)
+ raw_test.to_csv("predictions_test_TranslateLSTM_end_HEK293_"+output_col+".tsv",sep="\t",index=False)