From 5976cf31ea773b25b662da18f81cd35b7910e5ae Mon Sep 17 00:00:00 2001 From: mufeili Date: Fri, 25 Dec 2020 02:14:01 +0800 Subject: [PATCH] Update --- examples/reaction_prediction/rexgen_direct/README.md | 4 ++-- examples/reaction_prediction/rexgen_direct/clean.sh | 11 +++++++++++ 2 files changed, 13 insertions(+), 2 deletions(-) create mode 100644 examples/reaction_prediction/rexgen_direct/clean.sh diff --git a/examples/reaction_prediction/rexgen_direct/README.md b/examples/reaction_prediction/rexgen_direct/README.md index 3ae27170..34a02324 100644 --- a/examples/reaction_prediction/rexgen_direct/README.md +++ b/examples/reaction_prediction/rexgen_direct/README.md @@ -153,7 +153,6 @@ In addition, atom mapping information is provided. **Notice** - The atom mapping numbers in the rxn should be consecutive integers starting from 1, or it will report [the molAtomMapNumber issue](https://github.com/awslabs/dgl-lifesci/issues/33). To avoid the problem, you could convert the raw rxn smiles with explicit hydrgogen atoms to the rxn smiles without hydrogen atoms by RDKit befor adding map ids for the rxn smiles. @@ -210,7 +209,8 @@ Then, add map for the rxn smiles. ``` the oxygen atom will be labelled as 15. - +Lastly, if you use the scripts for multiple different datasets, you need to delete local cached files with `. clean.sh` +before working on a new dataset. You can then train a model on new datasets with diff --git a/examples/reaction_prediction/rexgen_direct/clean.sh b/examples/reaction_prediction/rexgen_direct/clean.sh new file mode 100644 index 00000000..658650f5 --- /dev/null +++ b/examples/reaction_prediction/rexgen_direct/clean.sh @@ -0,0 +1,11 @@ +rm -rf center_results +rm -rf candidate_results +rm -f train.bin +rm -f train_valid_reactions.proc +rm -f train_invalid_reactions.proc +rm -f val.bin +rm -f val_valid_reactions.proc +rm -f val_invalid_reactions.proc +rm -f test.bin +rm -f test_valid_reactions.proc +rm -f test_invalid_reactions.proc \ No newline at end of file