Microbiome在2018年发表了一篇基于深度学习鉴定抗性基因的工具DeepARG（https://microbiomejournal.biomedcentral.com/track/pdf/10.1186/s40168-018-0401-z?site=microbiomejournal.biomedcentral.com）

按照https://bitbucket.org/gusphdproj/deeparg-ss上的说明，对DeepARG在Linux系统上进行了安装

利用python的pip 进行安装：

1）首先运行 pip install nolearn

成功安装，提示信息为

“Installing collected packages: tabulate, numpy, Lasagne, scipy, six, Theano, gnumpy, gdbn, scikit-learn, joblib, nolearn
Successfully installed Lasagne-0.1 Theano-1.0.1 gdbn-0.1 gnumpy-0.2 joblib-0.11 nolearn-0.6.0 numpy-1.14.1 scikit-learn-0.19.1 scipy-1.0.0 six-1.11.0 tabulate-0.8.2”

2）其次进行测试，python deepARG.py --help

出现如下报错信息：

Exception: Compilation failed (return status=1): /usr/bin/ld: ./libpython3.5m.a(abstract.o): relocation R_X86_64_32S against `_Py_NotImplementedStruct' can not be used when making a shared object; recompile with -fPIC. ./libpython3.5m.a: could not read symbols: Bad value. collect2: ld returned 1 exit status.

Google到解决方案是python安装时，参数没有设置完全https://github.com/aigamedev/scikit-neuralnetwork/issues/215

于是重新安装python，安装命令

./configure --prefix=./ --enable-shared --build=x86_64-unknown-linux-gnu

make

make install

3）重新运行python deepARG.py --help

出现如下报错信息：

from Theano.tensor.signal import downsample

ImportError: No module named Theano.tensor.signal

Google到解决方案，https://github.com/Theano/Theano/issues/4337

于是修改lasagne/layers/pool.py中的相关代码

4）重新运行python deepARG.py --help

报错提示参数没用对，报错信息是getopt.GetoptError: option --help not recognized

于是修改为python deepARG.py -h

显示如下