You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Could you either update the vendored indelpost to the latest release from St Judes or cherrypick this change in? We have a number of samples that reliable trip this particular indelpost bug, where it will correctly run ~10-15% of the time but fail the remainder of times. For example, for this one sample (which I'm happy to provide via email):
for i in $(seq 0 20); uvx --with-requirements $PWD/3rdparty/aldy/requirements.txt aldy genotype --profile illumina -g ugt1a1 --cn=1,1 --output run.$i.aldy --genome hg19 sample.bam
17 of the runs fail with:
🐿 Aldy v4.7 (Python 3.13.1 on macOS 15.3.1)
(c) 2016-2025 Aldy Authors. All rights reserved.
Free for non-commercial/academic use only.
Genotyping sample sample.bam...
Trace back error: 0.
ERROR: gene= ugt1a1, file= sample.bam
AttributeError("'NoneType' object has no attribute 'count'")
Traceback (most recent call last):
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/__main__.py", line 122, in main
_genotype(args.gene, output, args)
~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/__main__.py", line 443, in _genotype
run(None)
~~~^^^^^^
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/__main__.py", line 393, in run
_ = genotype(
gene_db=gene,
...<8 lines>...
**{k: v for k, v in params.items() if v is not None},
)
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/genotype.py", line 186, in genotype
sample = sam.Sample(gene, profile, sam_path, reference, debug)
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/sam.py", line 117, in __init__
norm, muts = self._load_sam(path, reference, debug)
~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/sam.py", line 168, in _load_sam
self._realign_indels(tmp, sam, reference)
~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^
File ".cache/uv/archive-v0/cV0mTG-moBE_x3g4pb_YH/lib/python3.13/site-packages/aldy/sam.py", line 499, in _realign_indels
valn = VariantAlignment( # type: ignore
v,
...<4 lines>...
exact_match_for_shiftable=exact_match_for_shiftable,
)
File "aldy/indelpost/varaln.pyx", line 169, in aldy.indelpost.varaln.VariantAlignment.__cinit__
File "aldy/indelpost/varaln.pyx", line 299, in aldy.indelpost.varaln.VariantAlignment._VariantAlignment__parse_pileup
File "aldy/indelpost/varaln.pyx", line 1165, in aldy.indelpost.varaln.grid_search
File "aldy/indelpost/pileup.pyx", line 788, in aldy.indelpost.pileup.retarget
File "aldy/indelpost/pileup.pyx", line 788, in aldy.indelpost.pileup.retarget
File "aldy/indelpost/pileup.pyx", line 659, in aldy.indelpost.pileup.retarget
AttributeError: 'NoneType' object has no attribute 'count'
and 3 create the expected UGT1A1 result:
🐿 Aldy v4.7 (Python 3.13.1 on macOS 15.3.1)
(c) 2016-2025 Aldy Authors. All rights reserved.
Free for non-commercial/academic use only.
Genotyping sample XH2W3GDMX2_CHE-003710-A7_CGCTGTTGAC-CAAGAATTGG.bam...
Potential UGT1A1 gene structures for XH2W3GDMX2_CHE-003710-A7_CGCTGTTGAC-CAAGAATTGG:
1: 2x*1 (confidence: 100%)
Potential major UGT1A1 star-alleles for XH2W3GDMX2_CHE-003710-A7_CGCTGTTGAC-CAAGAATTGG:
1: 2x*1 (confidence: 100%)
Best UGT1A1 star-alleles for XH2W3GDMX2_CHE-003710-A7_CGCTGTTGAC-CAAGAATTGG:
1: *1 / *1 (confidence=100%)
Minor alleles: *1, *1
UGT1A1 results:
- *1 / *1
Minor: [*1] / [*1]
Legacy notation: [*1] / [*1]
Estimated activity for *1: unknown
Estimated activity for *1: unknown
The text was updated successfully, but these errors were encountered:
aldy/aldy/indelpost/pileup.pyx
Lines 653 to 656 in 0849a25
Hi @inumanag
There's a known bug in this indelpost function that was fixed/worked around about 2 years ago: https://github.com/stjude/indelPost/blob/ecad282eada13e6bb7e54c4b6fe5d32808cca233/indelpost/pileup.pyx#L654-L656
Could you either update the vendored indelpost to the latest release from St Judes or cherrypick this change in? We have a number of samples that reliable trip this particular indelpost bug, where it will correctly run ~10-15% of the time but fail the remainder of times. For example, for this one sample (which I'm happy to provide via email):
17 of the runs fail with:
and 3 create the expected UGT1A1 result:
The text was updated successfully, but these errors were encountered: