Error with mikado serialise: 'Engine' object has no attribute 'cursor'

[bbista]

Hello,
I am having trouble with the mikado serialize step. I get this error every time:

mikado serialise --json-conf configuration.yaml --xml mikado_prepared.blast.tsv --orfs mikado.orfs.gff3 --junctions portcullis_filtered.pass.junctions.bed -bt /uniprot/TSEprotein.faa

Mikado crashed, cause:
'Engine' object has no attribute 'cursor'
Traceback (most recent call last):
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/main.py", line 68, in main
args.func(args)
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/subprograms/serialise.py", line 383, in serialise
load_orfs(mikado_configuration, logger)
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/subprograms/serialise.py", line 183, in load_orfs
serializer()
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/serializers/orf.py", line 486, in call
self.serialize()
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/serializers/orf.py", line 461, in serialize
self.load_fasta()
File "/home/bbista/.local/lib/python3.9/site-packages/Mikado/serializers/orf.py", line 260, in load_fasta
cache = pd.read_sql_table("query", self.engine, index_col="query_name", columns=["query_name", "query_id"])
File "/home/bbista/.local/lib/python3.9/site-packages/pandas/io/sql.py", line 385, in read_sql_table
if not pandas_sql.has_table(table_name):
File "/home/bbista/.local/lib/python3.9/site-packages/pandas/io/sql.py", line 2863, in has_table
return len(self.execute(query, [name]).fetchall()) > 0
File "/home/bbista/.local/lib/python3.9/site-packages/pandas/io/sql.py", line 2670, in execute
cur = self.con.cursor()
AttributeError: 'Engine' object has no attribute 'cursor'

The log file for serialize looks like this:
2024-02-26 07:55:35,513 - serialise - serialise.py:321 - INFO - setup - MainProcess - Mikado version: 2.3.4
2024-02-26 07:55:35,513 - serialise - serialise.py:322 - INFO - setup - MainProcess - Command line: /home/bbista/.local/bin/mikado serialise --json-conf configuration.yaml --xml mikado_prepared.blast.tsv --orfs mikado.orfs.gff3 --junctions /work/LAS/nvalenzu-lab/bbista/ASP_final_annot/portcullis_out_ASP/3-filt/portcullis_filtered.pass.junctions.bed -bt /work/LAS/nvalenzu-lab/bbista/uniprot/TSEprotein.faa
2024-02-26 07:55:35,526 - serialise - serialise.py:332 - INFO - setup - MainProcess - Random seed: 0
2024-02-26 07:55:35,526 - serialise - serialise.py:345 - INFO - setup - MainProcess - Using a sqlite database (location: mikado.db)
2024-02-26 07:55:35,526 - serialise - serialise.py:348 - INFO - setup - MainProcess - Requested 1 threads, forcing single thread: False
2024-02-26 07:55:35,526 - serialise - serialise.py:176 - INFO - load_orfs - MainProcess - Starting to load ORF data

I am using Mikado v2.3.4 installed using the following steps:

mamba env create -f environment.yml

pip wheel -w dist .
pip install dist/*whl

I only have issues with the serialise part.

Any help is greatly appreaciated.

Best,
bbista

[jasonshiller]

@bbista I have the same problem. Did you manage to solve this?

[bbista]

@jasonshiller I did not solve it at all. I work in an HPC cluster and I just installed mikado in a different HPC cluster where I do not get this error. Only advice I can give you is to try a different machine.

[VitaxZhang]

I encountered the same problem as @jasonshiller and @bbista. I installed Mikado using Conda, and the version of Pandas provided was 2.2.x, while the version of SQLAlchemy was 1.4.x. The incompatibility between these two versions is documented pandas-dev/pandas#57049. I downgraded Pandas, and Mikado ran successfully afterward.

[gemygk]

Apologies @bbista @jasonshiller @VitaxZhang

I wanted to let you know that I added additional methods for installing Mikado (via Docker, Singularity, and Conda/Mamba) a couple of months ago. Please refer to the updated installation steps here for detailed installation steps.

Please let me know how it goes.