转座子分类软件deepTE使用中遇到的报错

deepTE的github主页

https://github.com/LiLabAtVT/DeepTE

单独新建一个虚拟环境

conda create -n DeepTE

安装3.6的python

conda activate DeepTE
conda install python=3.6 -y

然后安装指定的依赖python模块，这里python模块使用pip安装

pip install biopython==1.72-i  https://mirrors.aliyun.com/pypi/simple/
pip install keras==2.2.4 -i  https://mirrors.aliyun.com/pypi/simple/
pip install tensorflow==1.14.0 -i  https://mirrors.aliyun.com/pypi/simple/
pip install sklearn -i  https://mirrors.aliyun.com/pypi/simple/

然后使用conda安装hmmer

conda install hmmer -y

https://de.cyverse.org/dl/d/89D2FE7A-41BA-4F64-80E2-B9C26D49E99F/Plants_model.tar.gz

把这个文件下载下来再解压出来

运行命令

python /home/myan/biotools/DeepTE/DeepTE.py -i /home/myan/my_data/raw_data/longan/reference/longan-families_unknown.fa -sp P -d working.dir -o output_dir02 -m_dir /home/myan/biotools/DeepTE/Plants_model

会有报错

  File "/home/myan/anaconda3/envs/DeepTE/lib/python3.6/site-packages/keras/engine/saving.py", line 224, in _deserialize_model
    model_config = json.loads(model_config.decode('utf-8'))

AttributeError: 'str' object has no attribute 'decode'

我这里的办法是把saving.py这个文件里的所有.decode('utf-8')都给删掉

sed -i "s/.decode('utf-8')//g" /home/myan/anaconda3/envs/DeepTE/lib/python3.6/site-packages/keras/engine/saving.py

接下来还有报错

  File "/home/myan/anaconda3/envs/DeepTE/lib/python3.6/site-packages/keras/engine/saving.py", line 229, in _deserialize_model
    original_keras_version = model_weights_group['keras_version'].decode('utf8')
AttributeError: 'str' object has no attribute 'decode'

再把 把saving.py这个文件里的所有.decode('utf8')都给删掉

sed -i "s/.decode('utf8')//g" /home/myan/anaconda3/envs/DeepTE/lib/python3.6/site-packages/keras/engine/saving.py

这次再运行如下命令

python /home/myan/biotools/DeepTE/DeepTE.py -i /home/myan/my_data/raw_data/longan/reference/longan-families_unknown.fa -sp P -d working.dir -o output_dir02 -m_dir /home/myan/biotools/DeepTE/Plants_model

就没有报错了，

但是会有一堆提示信息

整个运行过程还挺快的，2000多条序列，90多秒就运行完了

输出文件夹 output_dir02必须提前新建好 working.dir这个文件夹会自己创建
output_dir02有两个文件

image.png image.png

会在原fasta文件的基础上添加分类信息

image.png

这个两列，第一列是原fasta的id 第二列是新的分类信息