From f6718da9c9b1fe0e23c48df1be3ea7775bdd92c2 Mon Sep 17 00:00:00 2001
From: korikuzma <korikuzma@gmail.com>
Date: Mon, 17 Oct 2022 17:43:06 -0400
Subject: [PATCH] build: separate out etl (dev) dependencies

---
 Pipfile                                       |  11 +-
 README.md                                     |  25 +-
 gene/etl/ensembl.py                           |   2 +-
 gene/etl/hgnc.py                              |  13 +-
 gene/etl/ncbi.py                              |  16 +-
 gene/{ => etl}/vrs_locations/__init__.py      |   0
 .../vrs_locations/chromosome_location.py      |   0
 .../vrs_locations/sequence_location.py        |  13 +-
 gene/version.py                               |   2 +-
 requirements-dev.txt                          | 264 +++++++++---------
 requirements.txt                              | 110 ++------
 setup.cfg                                     |  15 +-
 .../{ensembl_106.gff3 => ensembl_107.gff3}    |   0
 tests/unit/test_database_and_etl.py           |   2 +-
 tests/unit/test_ensembl_source.py             |   4 +-
 15 files changed, 214 insertions(+), 263 deletions(-)
 rename gene/{ => etl}/vrs_locations/__init__.py (100%)
 rename gene/{ => etl}/vrs_locations/chromosome_location.py (100%)
 rename gene/{ => etl}/vrs_locations/sequence_location.py (88%)
 rename tests/unit/data/etl_data/{ensembl_106.gff3 => ensembl_107.gff3} (100%)

diff --git a/Pipfile b/Pipfile
index 829cc80e..c7f0cf2e 100644
--- a/Pipfile
+++ b/Pipfile
@@ -9,15 +9,14 @@ fastapi = "*"
 uvicorn = "*"
 click = "*"
 boto3 = "*"
-beautifulsoup4 = "*"
-gffutils = "*"
-requests = "*"
-"biocommons.seqrepo" = "*"
-"ga4gh.vrs" = {version = ">=0.7.5.dev1", extras = ["extras"]}
-"ga4gh.vrsatile.pydantic" = ">=0.0.10"
+"ga4gh.vrsatile.pydantic" = "==0.0.11"
 
 [dev-packages]
 gene = {editable = true, path = "."}
+gffutils = "*"
+"biocommons.seqrepo" = "*"
+"ga4gh.vrs" = {version = "==0.8.0dev0", extras = ["extras"]}
+psycopg2-binary = "*"
 pytest = "*"
 pre-commit = "*"
 flake8 = "*"
diff --git a/README.md b/README.md
index 401b7ba7..b8256e7b 100644
--- a/README.md
+++ b/README.md
@@ -1,6 +1,14 @@
 # Gene Normalization
 Services and guidelines for normalizing gene terms
 
+Installing with pip:
+
+```commandline
+pip install gene[dev]
+```
+
+The `[dev]` argument tells pip to install packages to fulfill the dependencies of the `gene.etl` package.
+
 ## Developer instructions
 Following are sections include instructions specifically for developers.
 
@@ -14,16 +22,29 @@ Once installed, from the project root dir, just run:
 ```commandline
 pipenv shell
 pipenv lock && pipenv sync
+pipenv install --dev
 ```
 
-Gene Normalization relies on [SeqRepo](https://github.com/biocommons/biocommons.seqrepo) data.
+Gene Normalization relies on [SeqRepo](https://github.com/biocommons/biocommons.seqrepo) data, which you must download yourself.
 
 From the _root_ directory:
 ```
 pip install seqrepo
 sudo mkdir /usr/local/share/seqrepo
 sudo chown $USER /usr/local/share/seqrepo
-seqrepo pull -i 2021-01-29
+seqrepo pull -i 2021-01-29  # Replace with latest version using `seqrepo list-remote-instances` if outdated
+```
+
+If you get an error similar to the one below:
+```
+PermissionError: [Error 13] Permission denied: '/usr/local/share/seqrepo/2021-01-29._fkuefgd' -> '/usr/local/share/seqrepo/2021-01-29'
+```
+
+You will want to do the following:\
+(*Might not be ._fkuefgd, so replace with your error message path*)
+```console
+sudo mv /usr/local/share/seqrepo/2021-01-29._fkuefgd /usr/local/share/seqrepo/2021-01-29
+exit
 ```
 
 ### Deploying DynamoDB Locally
diff --git a/gene/etl/ensembl.py b/gene/etl/ensembl.py
index 2ea25642..13720d5a 100644
--- a/gene/etl/ensembl.py
+++ b/gene/etl/ensembl.py
@@ -9,7 +9,7 @@
 from gene import APP_ROOT
 from gene.schemas import SourceName, NamespacePrefix, Strand, SourceMeta
 from gene.database import Database
-from gene.vrs_locations import SequenceLocation
+from gene.etl.vrs_locations import SequenceLocation
 
 
 logger = logging.getLogger("gene")
diff --git a/gene/etl/hgnc.py b/gene/etl/hgnc.py
index 98cd2e6b..205c0f85 100644
--- a/gene/etl/hgnc.py
+++ b/gene/etl/hgnc.py
@@ -1,14 +1,15 @@
 """This module defines the HGNC ETL methods."""
-from .base import Base
-from gene import APP_ROOT, PREFIX_LOOKUP
-from gene.schemas import SourceName, SymbolStatus, NamespacePrefix, \
-    SourceMeta, Annotation, Chromosome
-from gene.database import Database
 import logging
 import json
 import shutil
 import re
-from gene.vrs_locations import ChromosomeLocation
+
+from gene import APP_ROOT, PREFIX_LOOKUP
+from gene.database import Database
+from gene.schemas import SourceName, SymbolStatus, NamespacePrefix, \
+    SourceMeta, Annotation, Chromosome
+from gene.etl.base import Base
+from gene.etl.vrs_locations import ChromosomeLocation
 
 logger = logging.getLogger('gene')
 logger.setLevel(logging.DEBUG)
diff --git a/gene/etl/ncbi.py b/gene/etl/ncbi.py
index e1a88062..e1239f2f 100644
--- a/gene/etl/ncbi.py
+++ b/gene/etl/ncbi.py
@@ -1,17 +1,19 @@
 """This module defines ETL methods for the NCBI data source."""
-from .base import Base
-from gene import APP_ROOT, PREFIX_LOOKUP
-from gene.database import Database
-from gene.schemas import SourceMeta, SourceName, NamespacePrefix, Annotation, \
-    Chromosome, SymbolStatus
+from ftplib import FTP
 import logging
 from pathlib import Path
 import csv
 from datetime import datetime
 import re
+
 import gffutils
-from gene.vrs_locations import SequenceLocation, ChromosomeLocation
-from ftplib import FTP
+
+from gene import APP_ROOT, PREFIX_LOOKUP
+from gene.database import Database
+from gene.schemas import SourceMeta, SourceName, NamespacePrefix, Annotation, \
+    Chromosome, SymbolStatus
+from gene.etl.base import Base
+from gene.etl.vrs_locations import SequenceLocation, ChromosomeLocation
 
 
 logger = logging.getLogger('gene')
diff --git a/gene/vrs_locations/__init__.py b/gene/etl/vrs_locations/__init__.py
similarity index 100%
rename from gene/vrs_locations/__init__.py
rename to gene/etl/vrs_locations/__init__.py
diff --git a/gene/vrs_locations/chromosome_location.py b/gene/etl/vrs_locations/chromosome_location.py
similarity index 100%
rename from gene/vrs_locations/chromosome_location.py
rename to gene/etl/vrs_locations/chromosome_location.py
diff --git a/gene/vrs_locations/sequence_location.py b/gene/etl/vrs_locations/sequence_location.py
similarity index 88%
rename from gene/vrs_locations/sequence_location.py
rename to gene/etl/vrs_locations/sequence_location.py
index 2dcc8f3e..d33ecd11 100644
--- a/gene/vrs_locations/sequence_location.py
+++ b/gene/etl/vrs_locations/sequence_location.py
@@ -1,8 +1,9 @@
 """This module defines GA4GH Sequence Location."""
 from typing import List
+import logging
+
 from ga4gh.vrs import models
 from ga4gh.core import ga4gh_identify
-import logging
 
 logger = logging.getLogger('gene')
 logger.setLevel(logging.DEBUG)
@@ -18,7 +19,12 @@ def get_aliases(self, sr, seqid) -> List[str]:
         :param str seqid: Sequence ID accession
         :return: List of aliases for seqid
         """
-        return sr.translate_alias(seqid)
+        aliases = []
+        try:
+            aliases = sr.translate_alias(seqid)
+        except KeyError as e:
+            logger.warning(f"SeqRepo raised KeyError: {e}")
+        return aliases
 
     def add_location(self, seqid, gene, params, sr):
         """Get a gene's Sequence Location.
@@ -31,6 +37,9 @@ def add_location(self, seqid, gene, params, sr):
         """
         location = dict()
         aliases = self.get_aliases(sr, seqid)
+        if not aliases:
+            return location
+
         sequence_id = [a for a in aliases if a.startswith('ga4gh')][0]
 
         if gene.start != '.' and gene.end != '.' and sequence_id:
diff --git a/gene/version.py b/gene/version.py
index 3fd08450..7f36ece5 100644
--- a/gene/version.py
+++ b/gene/version.py
@@ -1,2 +1,2 @@
 """Gene normalizer version"""
-__version__ = "0.1.27"
+__version__ = "0.1.28"
diff --git a/requirements-dev.txt b/requirements-dev.txt
index 93f0cabf..fbc0854f 100644
--- a/requirements-dev.txt
+++ b/requirements-dev.txt
@@ -1,164 +1,156 @@
-#
-# These requirements were autogenerated by pipenv
-# To regenerate from the project's Pipfile, run:
-#
-#    pipenv lock --requirements --dev
-#
-
-# Note: in pipenv 2020.x, "--dev" changed to emit both default and development
-# requirements. To emit only development requirements, pass "--dev-only".
-
 -i https://pypi.org/simple
--e .
-anyio==3.5.0; python_full_version >= '3.6.2'
+anyio==3.6.1 ; python_full_version >= '3.6.2'
 appdirs==1.4.4
-appnope==0.1.2; sys_platform == 'darwin'
-argcomplete==2.0.0; python_version >= '3.6'
+appnope==0.1.3 ; sys_platform == 'darwin'
+argcomplete==2.0.0 ; python_version >= '3.6'
 argh==0.26.2
-argon2-cffi-bindings==21.2.0; python_version >= '3.6'
-argon2-cffi==21.3.0; python_version >= '3.6'
-asgiref==3.5.0; python_version >= '3.7'
-asttokens==2.0.5
-attrs==21.4.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-babel==2.9.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+argon2-cffi==21.3.0 ; python_version >= '3.6'
+argon2-cffi-bindings==21.2.0 ; python_version >= '3.6'
+asttokens==2.0.8
+attrs==22.1.0 ; python_version >= '3.5'
+babel==2.10.3 ; python_version >= '3.6'
 backcall==0.2.0
-beautifulsoup4==4.10.0
+beautifulsoup4==4.11.1 ; python_full_version >= '3.6.0'
 biocommons.seqrepo==0.6.5
-bioutils==0.5.5; python_version >= '3.6'
-bleach==4.1.0; python_version >= '3.6'
-boto3==1.21.26
-botocore==1.24.26; python_version >= '3.6'
+bioutils==0.5.7 ; python_version >= '3.6'
+bleach==5.0.1 ; python_version >= '3.7'
+boto3==1.24.92
+botocore==1.27.92 ; python_version >= '3.7'
 bs4==0.0.1
-canonicaljson==1.6.0; python_version ~= '3.7'
-certifi==2021.10.8
-cffi==1.15.0
-cfgv==3.3.1; python_full_version >= '3.6.1'
-charset-normalizer==2.0.12; python_version >= '3'
-click==8.0.4
-coloredlogs==15.0.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-configparser==5.2.0; python_version >= '3.6'
-coverage==6.3.2
+canonicaljson==1.6.3 ; python_version >= '3.7'
+certifi==2022.9.24 ; python_version >= '3.6'
+cffi==1.15.1
+cfgv==3.3.1 ; python_full_version >= '3.6.1'
+charset-normalizer==2.1.1 ; python_full_version >= '3.6.0'
+click==8.1.3
+coloredlogs==15.0.1 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+configparser==5.3.0 ; python_version >= '3.7'
+coverage==6.5.0
 coveralls==3.3.1
-cssselect==1.1.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-debugpy==1.6.0; python_version >= '3.7'
-decorator==5.1.1; python_version >= '3.5'
-defusedxml==0.7.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-distlib==0.3.4
+cssselect==1.1.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+debugpy==1.6.3 ; python_version >= '3.7'
+decorator==5.1.1 ; python_version >= '3.5'
+defusedxml==0.7.1 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+distlib==0.3.6
 docopt==0.6.2
-entrypoints==0.4; python_version >= '3.6'
-executing==0.8.3
+entrypoints==0.4 ; python_version >= '3.6'
+executing==1.1.1
 fake-useragent==0.1.11
-fastapi==0.75.0
-filelock==3.6.0; python_version >= '3.7'
+fastapi==0.85.1
+fastjsonschema==2.16.2
+filelock==3.8.0 ; python_version >= '3.7'
+flake8==5.0.4
 flake8-docstrings==1.6.0
-flake8==4.0.1
-ga4gh.vrs[extras]==0.8.0.dev0
-ga4gh.vrsatile.pydantic==0.0.10
-gffutils==0.10.1
-h11==0.13.0; python_version >= '3.6'
+ga4gh.vrs[extras]==0.8.0dev0
+ga4gh.vrsatile.pydantic==0.0.11
+-e .
+gffutils==0.11.1
+h11==0.14.0 ; python_version >= '3.7'
 hgvs==1.5.2
-humanfriendly==10.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-identify==2.4.12; python_version >= '3.7'
-idna==3.3; python_version >= '3'
-importlib-metadata==4.11.3; python_version < '3.10'
-inflection==0.5.1; python_version >= '3.5'
+humanfriendly==10.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+identify==2.5.6 ; python_version >= '3.7'
+idna==3.4 ; python_version >= '3.5'
+importlib-metadata==5.0.0 ; python_version >= '3.7'
+inflection==0.5.1 ; python_version >= '3.5'
 iniconfig==1.1.1
-ipykernel==6.9.2; python_version >= '3.7'
+ipykernel==6.16.0 ; python_version >= '3.7'
+ipython==8.5.0 ; python_version >= '3.8'
 ipython-genutils==0.2.0
-ipython==8.1.1; python_version >= '3.8'
-jedi==0.18.1; python_version >= '3.6'
-jinja2==3.1.0; python_version >= '3.7'
-jmespath==1.0.0; python_version >= '3.7'
-json5==0.9.6
+jedi==0.18.1 ; python_version >= '3.6'
+jinja2==3.1.2 ; python_version >= '3.7'
+jmespath==1.0.1 ; python_version >= '3.7'
+json5==0.9.10
 jsonschema==3.2.0
-jupyter-client==7.1.2; python_full_version >= '3.6.1'
-jupyter-core==4.9.2; python_version >= '3.6'
-jupyter-server==1.15.6; python_version >= '3.7'
-jupyterlab-pygments==0.1.2
-jupyterlab-server==2.11.2; python_version >= '3.7'
-jupyterlab==3.3.2
-lxml==4.8.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-markdown==3.3.6; python_version >= '3.6'
-markupsafe==2.1.1; python_version >= '3.7'
-matplotlib-inline==0.1.3; python_version >= '3.5'
-mccabe==0.6.1
-mistune==0.8.4
+jupyter-client==7.4.2 ; python_version >= '3.7'
+jupyter-core==4.11.1 ; python_version >= '3.7'
+jupyter-server==1.21.0 ; python_version >= '3.7'
+jupyterlab==3.4.8
+jupyterlab-pygments==0.2.2 ; python_version >= '3.7'
+jupyterlab-server==2.16.0 ; python_version >= '3.7'
+lxml==4.9.1 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+markdown==3.4.1 ; python_version >= '3.7'
+markupsafe==2.1.1 ; python_version >= '3.7'
+matplotlib-inline==0.1.6 ; python_version >= '3.5'
+mccabe==0.7.0 ; python_version >= '3.6'
+mistune==2.0.4
 mock==4.0.3
-nbclassic==0.3.7; python_version >= '3.7'
-nbclient==0.5.13; python_version >= '3.7'
-nbconvert==6.4.4; python_version >= '3.7'
-nbformat==5.2.0; python_version >= '3.7'
-nest-asyncio==1.5.4; python_version >= '3.5'
-nodeenv==1.6.0
-notebook-shim==0.1.0; python_version >= '3.7'
-notebook==6.4.10; python_version >= '3.6'
-numpy==1.22.3; python_version >= '3.8'
-packaging==21.3; python_version >= '3.6'
-pandocfilters==1.5.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+nbclassic==0.4.5 ; python_version >= '3.7'
+nbclient==0.7.0 ; python_full_version >= '3.7.0'
+nbconvert==7.2.1 ; python_version >= '3.7'
+nbformat==5.7.0 ; python_version >= '3.7'
+nest-asyncio==1.5.6 ; python_version >= '3.5'
+nodeenv==1.7.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4, 3.5, 3.6'
+notebook==6.5.1 ; python_version >= '3.7'
+notebook-shim==0.2.0 ; python_version >= '3.7'
+numpy==1.23.4 ; python_version >= '3.8'
+packaging==21.3 ; python_version >= '3.6'
+pandocfilters==1.5.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
 parse==1.19.0
 parsley==1.3
-parso==0.8.3; python_version >= '3.6'
-pexpect==4.8.0; sys_platform != 'win32'
+parso==0.8.3 ; python_version >= '3.6'
+pexpect==4.8.0 ; sys_platform != 'win32'
 pickleshare==0.7.5
-platformdirs==2.5.1; python_version >= '3.7'
-pluggy==1.0.0; python_version >= '3.6'
-pre-commit==2.17.0
-prometheus-client==0.13.1; python_version >= '3.6'
-prompt-toolkit==3.0.28; python_full_version >= '3.6.2'
-psutil==5.9.0; python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'
-psycopg2==2.9.3; python_version >= '3.6'
+platformdirs==2.5.2 ; python_version >= '3.7'
+pluggy==1.0.0 ; python_version >= '3.6'
+pre-commit==2.20.0
+prometheus-client==0.15.0 ; python_version >= '3.6'
+prompt-toolkit==3.0.31 ; python_full_version >= '3.6.2'
+psutil==5.9.2 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+psycopg2==2.9.4 ; python_version >= '3.6'
+psycopg2-binary==2.9.4
 ptyprocess==0.7.0
 pure-eval==0.2.2
-py==1.11.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-pycodestyle==2.8.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+py==1.11.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
+pycodestyle==2.9.1 ; python_version >= '3.6'
 pycparser==2.21
-pydantic==1.9.0
-pydocstyle==6.1.1; python_version >= '3.6'
+pydantic==1.10.2
+pydocstyle==6.1.1 ; python_version >= '3.6'
 pyee==8.2.2
-pyfaidx==0.6.4
-pyflakes==2.4.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-pygments==2.11.2; python_version >= '3.5'
-pyparsing==3.0.7; python_version >= '3.6'
-pyppeteer==1.0.2; python_version >= '3.7' and python_version < '4.0'
+pyfaidx==0.7.1
+pyflakes==2.5.0 ; python_version >= '3.6'
+pygments==2.13.0 ; python_version >= '3.6'
+pyparsing==3.0.9 ; python_full_version >= '3.6.8'
+pyppeteer==1.0.2 ; python_version >= '3.7' and python_version < '4.0'
 pyquery==1.4.3
-pyrsistent==0.18.1; python_version >= '3.7'
-pysam==0.18.0
-pytest-cov==3.0.0
-pytest==7.1.1
-python-dateutil==2.8.2; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+pyrsistent==0.18.1 ; python_version >= '3.7'
+pysam==0.19.1
+pytest==7.1.3
+pytest-cov==4.0.0
+python-dateutil==2.8.2 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
 python-jsonschema-objects==0.4.1
-pytz==2022.1
-pyyaml==6.0; python_version >= '3.6'
-pyzmq==22.3.0; python_version >= '3.6'
-requests-html==0.10.0; python_version >= '3.6'
-requests==2.27.1
-s3transfer==0.5.2; python_version >= '3.6'
+pytz==2022.4
+pyyaml==6.0 ; python_version >= '3.6'
+pyzmq==24.0.1 ; python_version >= '3.6'
+requests==2.28.1
+requests-html==0.10.0 ; python_full_version >= '3.6.0'
+s3transfer==0.6.0 ; python_version >= '3.7'
 send2trash==1.8.0
-setuptools==61.0.0; python_version >= '3.7'
-simplejson==3.17.6; python_version >= '2.5' and python_version not in '3.0, 3.1, 3.2, 3.3'
-six==1.16.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-sniffio==1.2.0; python_version >= '3.5'
+setuptools==65.5.0 ; python_version >= '3.7'
+simplejson==3.17.6 ; python_version >= '2.5' and python_version not in '3.0, 3.1, 3.2, 3.3'
+six==1.16.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+sniffio==1.3.0 ; python_version >= '3.7'
 snowballstemmer==2.2.0
-soupsieve==2.3.1; python_version >= '3.6'
-sqlparse==0.4.2; python_version >= '3.5'
-stack-data==0.2.0
-starlette==0.17.1; python_version >= '3.6'
-tabulate==0.8.9
-terminado==0.13.3; python_version >= '3.7'
-testpath==0.6.0; python_version >= '3.5'
-toml==0.10.2; python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'
-tomli==2.0.1; python_version >= '3.7'
-tornado==6.1; python_version >= '3.5'
-tqdm==4.63.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-traitlets==5.1.1; python_version >= '3.7'
-typing-extensions==4.1.1; python_version >= '3.6'
-urllib3==1.26.9; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4' and python_version < '4.0'
-uvicorn==0.17.6
-virtualenv==20.13.4; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-w3lib==1.22.0
+soupsieve==2.3.2.post1 ; python_version >= '3.6'
+sqlparse==0.4.3 ; python_version >= '3.5'
+stack-data==0.5.1
+starlette==0.20.4 ; python_version >= '3.7'
+tabulate==0.9.0 ; python_version >= '3.7'
+terminado==0.16.0 ; python_version >= '3.7'
+tinycss2==1.2.0 ; python_version >= '3.7'
+toml==0.10.2 ; python_version >= '2.6' and python_version not in '3.0, 3.1, 3.2, 3.3'
+tomli==2.0.1 ; python_version >= '3.7'
+tornado==6.2 ; python_version >= '3.7'
+tqdm==4.64.1 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+traitlets==5.4.0 ; python_version >= '3.7'
+typing-extensions==4.4.0 ; python_version >= '3.7'
+urllib3==1.26.12 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4, 3.5' and python_version < '4'
+uvicorn==0.18.3
+virtualenv==20.16.5 ; python_version >= '3.6'
+w3lib==2.0.1 ; python_version >= '3.6'
 wcwidth==0.2.5
 webencodings==0.5.1
-websocket-client==1.3.1; python_version >= '3.6'
-websockets==10.2; python_version >= '3.7'
-yoyo-migrations==7.3.2
-zipp==3.7.0; python_version >= '3.7'
+websocket-client==1.4.1 ; python_version >= '3.7'
+websockets==10.3 ; python_version >= '3.7'
+-e .
+yoyo-migrations==8.0.0
+zipp==3.9.0 ; python_version >= '3.7'
diff --git a/requirements.txt b/requirements.txt
index 7386be8f..bcf61a84 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -1,93 +1,19 @@
-#
-# These requirements were autogenerated by pipenv
-# To regenerate from the project's Pipfile, run:
-#
-#    pipenv lock --requirements
-#
-
 -i https://pypi.org/simple
-anyio==3.5.0; python_full_version >= '3.6.2'
-appdirs==1.4.4
-appnope==0.1.2; sys_platform == 'darwin'
-argcomplete==2.0.0; python_version >= '3.6'
-argh==0.26.2
-asgiref==3.5.0; python_version >= '3.7'
-asttokens==2.0.5
-attrs==21.4.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-backcall==0.2.0
-beautifulsoup4==4.10.0
-biocommons.seqrepo==0.6.5
-bioutils==0.5.5; python_version >= '3.6'
-boto3==1.21.26
-botocore==1.24.26; python_version >= '3.6'
-bs4==0.0.1
-canonicaljson==1.6.0; python_version ~= '3.7'
-certifi==2021.10.8
-charset-normalizer==2.0.12; python_version >= '3'
-click==8.0.4
-coloredlogs==15.0.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-configparser==5.2.0; python_version >= '3.6'
-cssselect==1.1.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-decorator==5.1.1; python_version >= '3.5'
-executing==0.8.3
-fake-useragent==0.1.11
-fastapi==0.75.0
-ga4gh.vrs[extras]==0.8.0.dev0
-ga4gh.vrsatile.pydantic==0.0.10
-gffutils==0.10.1
-h11==0.13.0; python_version >= '3.6'
-hgvs==1.5.2
-humanfriendly==10.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-idna==3.3; python_version >= '3'
-importlib-metadata==4.11.3; python_version < '3.10'
-inflection==0.5.1; python_version >= '3.5'
-ipython==8.1.1; python_version >= '3.8'
-jedi==0.18.1; python_version >= '3.6'
-jmespath==1.0.0; python_version >= '3.7'
-jsonschema==3.2.0
-lxml==4.8.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4'
-markdown==3.3.6; python_version >= '3.6'
-matplotlib-inline==0.1.3; python_version >= '3.5'
-numpy==1.22.3; python_version >= '3.8'
-parse==1.19.0
-parsley==1.3
-parso==0.8.3; python_version >= '3.6'
-pexpect==4.8.0; sys_platform != 'win32'
-pickleshare==0.7.5
-prompt-toolkit==3.0.28; python_full_version >= '3.6.2'
-psycopg2==2.9.3; python_version >= '3.6'
-ptyprocess==0.7.0
-pure-eval==0.2.2
-pydantic==1.9.0
-pyee==8.2.2
-pyfaidx==0.6.4
-pygments==2.11.2; python_version >= '3.5'
-pyppeteer==1.0.2; python_version >= '3.7' and python_version < '4.0'
-pyquery==1.4.3
-pyrsistent==0.18.1; python_version >= '3.7'
-pysam==0.18.0
-python-dateutil==2.8.2; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-python-jsonschema-objects==0.4.1
-pyyaml==6.0; python_version >= '3.6'
-requests-html==0.10.0; python_version >= '3.6'
-requests==2.27.1
-s3transfer==0.5.2; python_version >= '3.6'
-setuptools==61.0.0; python_version >= '3.7'
-simplejson==3.17.6; python_version >= '2.5' and python_version not in '3.0, 3.1, 3.2, 3.3'
-six==1.16.0; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-sniffio==1.2.0; python_version >= '3.5'
-soupsieve==2.3.1; python_version >= '3.6'
-sqlparse==0.4.2; python_version >= '3.5'
-stack-data==0.2.0
-starlette==0.17.1; python_version >= '3.6'
-tabulate==0.8.9
-tqdm==4.63.1; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
-traitlets==5.1.1; python_version >= '3.7'
-typing-extensions==4.1.1; python_version >= '3.6'
-urllib3==1.26.9; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4' and python_version < '4.0'
-uvicorn==0.17.6
-w3lib==1.22.0
-wcwidth==0.2.5
-websockets==10.2; python_version >= '3.7'
-yoyo-migrations==7.3.2
-zipp==3.7.0; python_version >= '3.7'
+anyio==3.6.1 ; python_full_version >= '3.6.2'
+boto3==1.24.92
+botocore==1.27.92 ; python_version >= '3.7'
+click==8.1.3
+fastapi==0.85.1
+ga4gh.vrsatile.pydantic==0.0.11
+h11==0.14.0 ; python_version >= '3.7'
+idna==3.4 ; python_version >= '3.5'
+jmespath==1.0.1 ; python_version >= '3.7'
+pydantic==1.10.2
+python-dateutil==2.8.2 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+s3transfer==0.6.0 ; python_version >= '3.7'
+six==1.16.0 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3'
+sniffio==1.3.0 ; python_version >= '3.7'
+starlette==0.20.4 ; python_version >= '3.7'
+typing-extensions==4.4.0 ; python_version >= '3.7'
+urllib3==1.26.12 ; python_version >= '2.7' and python_version not in '3.0, 3.1, 3.2, 3.3, 3.4, 3.5' and python_version < '4'
+uvicorn==0.18.3
diff --git a/setup.cfg b/setup.cfg
index 1dfcc3c0..9b99d78c 100644
--- a/setup.cfg
+++ b/setup.cfg
@@ -29,18 +29,19 @@ install_requires =
     uvicorn
     click
     boto3
-    beautifulsoup4
-    gffutils
-    requests
-    biocommons.seqrepo
-    ga4gh.vrs[extras] >= 0.7.5.dev1
-    ga4gh.vrsatile.pydantic >= 0.0.10
-
+    ga4gh.vrsatile.pydantic == 0.0.11
 
 tests_require =
     pytest
     pytest-cov
     mock
 
+[options.extras_require]
+dev =
+    gffutils
+    biocommons.seqrepo
+    ga4gh.vrs[extras] == 0.8.0dev0
+    psycopg2-binary
+
 [tool:pytest]
 addopts = --ignore setup.py --ignore codebuild/ --ignore dynamodb_local_latest/ --doctest-modules --cov-report term-missing --cov .
diff --git a/tests/unit/data/etl_data/ensembl_106.gff3 b/tests/unit/data/etl_data/ensembl_107.gff3
similarity index 100%
rename from tests/unit/data/etl_data/ensembl_106.gff3
rename to tests/unit/data/etl_data/ensembl_107.gff3
diff --git a/tests/unit/test_database_and_etl.py b/tests/unit/test_database_and_etl.py
index 25584a29..8b7d88fe 100644
--- a/tests/unit/test_database_and_etl.py
+++ b/tests/unit/test_database_and_etl.py
@@ -91,7 +91,7 @@ def test_ensembl_etl(test_get_seqrepo, processed_ids, dynamodb, etl_data_path,
         shutil.rmtree(e.src_data_dir)
 
         e._sequence_location.get_aliases = _get_aliases
-        e._data_src = etl_data_path / 'ensembl_106.gff3'
+        e._data_src = etl_data_path / 'ensembl_107.gff3'
         e._transform_data()
         e._add_meta()
         processed_ids += e._processed_ids
diff --git a/tests/unit/test_ensembl_source.py b/tests/unit/test_ensembl_source.py
index 907d88e4..8a27eb03 100644
--- a/tests/unit/test_ensembl_source.py
+++ b/tests/unit/test_ensembl_source.py
@@ -304,9 +304,9 @@ def test_meta_info(ensembl):
     assert resp.source_meta_.data_license == "custom"
     assert resp.source_meta_.data_license_url == \
         "https://useast.ensembl.org/info/about/legal/disclaimer.html"
-    assert resp.source_meta_.version == "106"
+    assert resp.source_meta_.version == "107"
     assert resp.source_meta_.data_url == \
-        "ftp://ftp.ensembl.org/pub/current_gff3/homo_sapiens/Homo_sapiens.GRCh38.106.gff3.gz"  # noqa: E501
+        "ftp://ftp.ensembl.org/pub/current_gff3/homo_sapiens/Homo_sapiens.GRCh38.107.gff3.gz"  # noqa: E501
     assert resp.source_meta_.rdp_url is None
     assert resp.source_meta_.genome_assemblies == ["GRCh38"]
     assert resp.source_meta_.data_license_attributes == {