Include non-python files

MANIFEST.in

@@ -0,0 +1,2 @@
+recursive-include data pseudoseqs.csv
+recursive-include models *

setup.py

@@ -12,6 +12,7 @@ def readme():
     url='https://github.com/KarchinLab/bigmhc',
     author='Albert, Benjamin Alexander and Yang, Yunxiao and Shao, Xiaoshan M. and Singh, Dipika and Smith, Kellie N. and Anagnostou, Valsamo and Karchin, Rachel',
     scripts=['src/bigmhc_predict', 'src/bigmhc_train'],
+    packages=['data', 'src', 'models'],
     include_package_data=True,
     install_requires=[
         'torch',

src/bigmhc_predict

@@ -17,9 +17,9 @@ import os
 import torch
 import pandas as pd
 
-import cli 
+import src.cli 
 
-from bigmhc import BigMHC
+from src.bigmhc import BigMHC
 
 
 def predict(models, data, args):
@@ -63,7 +63,7 @@ def predict(models, data, args):
 
 def main():
 
-    args, data, models = cli.parseArgs(train=False)
+    args, data, models = src.cli.parseArgs(train=False)
 
     preds = predict(models=models, data=data, args=args)
 

src/bigmhc_train

@@ -16,9 +16,9 @@ import os
 
 import torch
 
-import cli
+import src.cli
 
-from bigmhc import BigMHC
+from src.bigmhc import BigMHC
 
 
 def train(model, data, args):
@@ -81,7 +81,7 @@ def train(model, data, args):
 
 def main():
 
-    args, data, model = cli.parseArgs(train=True)
+    args, data, model = src.cli.parseArgs(train=True)
 
     if len(model) > 1:
         raise ValueError(

src/cli.py

@@ -28,10 +28,10 @@
 
 from typing import Union
 
-from bigmhc import BigMHC
+from src.bigmhc import BigMHC
 
-from dataset import Dataset
-from mhcenc import MHCEncoder
+from src.dataset import Dataset
+from src.mhcenc import MHCEncoder
 
 
 def _rootdir():

src/dataset.py

@@ -20,8 +20,8 @@
 
 from typing import Union, Callable
 
-import mhcuid
-import encseq
+import src.mhcuid
+import src.encseq
 
 
 class Batch:
@@ -52,7 +52,7 @@ def __init__(
             self,
             pmhcs  : pd.DataFrame,
             mhcenc : dict = None,
-            encpep : Callable = encseq.dummy):
+            encpep : Callable = src.encseq.dummy):
 
         self.df     = pmhcs
         self.mhcenc = mhcenc
@@ -280,7 +280,7 @@ def readPMHCs(
         df.insert(
             loc=0,
             column="uid",
-            value=df["mhc"].apply(mhcuid.mhcuid) + '_' + df["pep"])
+            value=df["mhc"].apply(src.mhcuid.mhcuid) + '_' + df["pep"])
         df.set_index(keys="uid", drop=True, inplace=True)
 
         if verbose:

src/mhcenc.py

@@ -37,7 +37,7 @@
 
 import torch
 
-from mhcuid import mhcuid
+from src.mhcuid import mhcuid
 
 
 class MHCEncoder: