strkit 0.25.0a2__tar.gz → 0.25.0a3__tar.gz

strkit-0.25.0a3/.gitignore

@@ -0,0 +1,43 @@
+/env
+/envp11
+__pycache__
+
+/build
+/dist
+/strkit.egg-info
+# ignore WIP cohort code for now
+/strkit/cohort
+# ignore WIP strkit viz code
+/strkit_viz
+
+plot-*.bash
+test-*.bash
+*.bam
+*.bai
+*.bcf
+*.dat
+*.fasta
+*.fa.gz
+*.fa.gz.fai
+*.fa.gz.gzi
+*.fastq
+*.fastq.gz
+*.fq
+*.fq.gz
+*.gff*
+*.gtf*
+*.log
+*.bed
+!tests/data/*.bed
+!tests/data/*.dat
+/*.json
+*.png
+*.time
+/*.tsv
+*.vcf.gz*
+*.vcf*
+
+*.token
+
+*.lprof
+*.poasta

{strkit-0.25.0a2 → strkit-0.25.0a3}/MANIFEST.in

@@ -1,4 +1,3 @@
-include strkit/VERSION
 include strkit/data/ploidy_configs/*.json
 include strkit/viz/static/logo.png
 include strkit/viz/templates/*.html

{strkit-0.25.0a2/strkit.egg-info → strkit-0.25.0a3}/PKG-INFO

@@ -1,43 +1,34 @@
 Metadata-Version: 2.4
 Name: strkit
-Version: 0.25.0a2
+Version: 0.25.0a3
 Summary: A toolkit for analyzing variation in short(ish) tandem repeats.
-Home-page: https://github.com/davidlougheed/strkit
-Author: David Lougheed
-Author-email: david.lougheed@gmail.com
-License: GPLv3
+Project-URL: Homepage, https://github.com/davidlougheed/strkit
+Project-URL: Repository, https://github.com/davidlougheed/strkit.git
+Author-email: David Lougheed <david.lougheed@gmail.com>
+License-Expression: GPL-3.0-only
+License-File: LICENSE
+Keywords: STRs,TRs,genomics,short tandem repeats,tandem repeats
+Classifier: Operating System :: MacOS :: MacOS X
+Classifier: Operating System :: POSIX
+Classifier: Operating System :: POSIX :: Linux
+Classifier: Programming Language :: Python :: 3 :: Only
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
 Classifier: Programming Language :: Python :: 3.12
 Classifier: Programming Language :: Python :: 3.13
-Classifier: Programming Language :: Python :: 3 :: Only
-Classifier: Operating System :: POSIX
-Classifier: Operating System :: POSIX :: Linux
-Classifier: Operating System :: MacOS :: MacOS X
-Requires-Python: ~=3.10
-Description-Content-Type: text/markdown
-License-File: LICENSE
-Requires-Dist: Flask<3.2,>=2.2.5
-Requires-Dist: orjson<4,>=3.11.1
-Requires-Dist: pysam<0.24,>=0.19
+Classifier: Programming Language :: Python :: 3.14
+Requires-Python: >=3.10
+Requires-Dist: flask<3.2,>=3.1
 Requires-Dist: numpy<3,>=2.1.3
+Requires-Dist: orjson<4,>=3.11.1
 Requires-Dist: parasail<1.4,>=1.2.4
 Requires-Dist: pydantic<3,>=2.11.7
-Requires-Dist: scikit-learn<1.8,>=1.2.1
+Requires-Dist: pysam<0.25,>=0.19
+Requires-Dist: scikit-learn<2,>=1.7.2
 Requires-Dist: scipy<1.17,>=1.10
 Requires-Dist: statsmodels<0.15,>=0.14.0
-Requires-Dist: strkit_rust_ext==0.26.0
-Dynamic: author
-Dynamic: author-email
-Dynamic: classifier
-Dynamic: description
-Dynamic: description-content-type
-Dynamic: home-page
-Dynamic: license
-Dynamic: license-file
-Dynamic: requires-dist
-Dynamic: requires-python
-Dynamic: summary
+Requires-Dist: strkit-rust-ext==0.28.0
+Description-Content-Type: text/markdown
 
 # STRkit - short tandem repeat genotyping with long reads
 
@@ -114,7 +105,7 @@ a long time (in the tens of minutes).
 On Digital Research Alliance of Canada/Compute Canada clusters, this involves loading a few modules:
 
 ```bash
-module load rust/1.91.0 clang/18.1.8 python/3.11 scipy-stack/2026a parasail/2.6.2
+module load rust/1.95.0 clang/18.1.8 python/3.13 scipy-stack/2026a parasail/2.6.2
 python -m pip install strkit
 ```
 
@@ -151,7 +142,7 @@ docker run \
   -it ghcr.io/davidlougheed/strkit call \
   /inputs/file.bam \
   --loci /inputs/loci.bed --ref /inputs/ref.fa --incorporate-snvs /inputs/dbsnp.vcf.gz \
-  --vcf /out/calls.vcf --no-tsv
+  --vcf /out/calls.vcf
 ```
 
 Inside the `inputs` directory, STRkit would (given these parameters) expect the following files to be present:
@@ -219,7 +210,6 @@ strkit call \
   --vcf my-calls.vcf \  # Calculate consensus sequences for alleles and output a .vcf (or .vcf.gz) with call data
   --seed 183 \  # Fixed random number generator seed for replicability
   --processes 10 \  # Number of parallel processes to use; DEFAULT: 1
-  --no-tsv  # If VCF output is enabled as above, we don't need TSV genotype output to stdout (which is the default)
 ```
 
 ##### REGARDING ALIGNMENTS
@@ -325,12 +315,12 @@ For more information, see also documentation on the [Output formats](./docs/outp
 
 If you want to output a full call report, you can use the `--json output-file.json` argument to
 specify a path to output a more detailed JSON document to. This document contains 99% CIs, peak
-labels, and some other information that isn't included in the normal TSV file. If you want this
+labels, and some other information that isn't included in the normal VCF file. If you want this
 file to be indented and human-readable, use the `--indent-json` flag in addition to `--json ...`.
 
-If you want to output a VCF file (STRs and SNVs if called; currently not phased), use the
-`--vcf ...` argument. If you pass `--vcf stdout`, the VCF will be written to `stdout` instead of a 
-file.
+VCF call information is printed to standard output by default (`--vcf stdout`). If you pass 
+`--vcf path/to/file.vcf`, the VCF will be written to a file instead. If you pass `--vcf none`, no
+VCF will be written.
 
 For more information, see also documentation on the [Output formats](./docs/output_formats.md).
 

{strkit-0.25.0a2 → strkit-0.25.0a3}/README.md

@@ -73,7 +73,7 @@ a long time (in the tens of minutes).
 On Digital Research Alliance of Canada/Compute Canada clusters, this involves loading a few modules:
 
 ```bash
-module load rust/1.91.0 clang/18.1.8 python/3.11 scipy-stack/2026a parasail/2.6.2
+module load rust/1.95.0 clang/18.1.8 python/3.13 scipy-stack/2026a parasail/2.6.2
 python -m pip install strkit
 ```
 
@@ -110,7 +110,7 @@ docker run \
   -it ghcr.io/davidlougheed/strkit call \
   /inputs/file.bam \
   --loci /inputs/loci.bed --ref /inputs/ref.fa --incorporate-snvs /inputs/dbsnp.vcf.gz \
-  --vcf /out/calls.vcf --no-tsv
+  --vcf /out/calls.vcf
 ```
 
 Inside the `inputs` directory, STRkit would (given these parameters) expect the following files to be present:
@@ -178,7 +178,6 @@ strkit call \
   --vcf my-calls.vcf \  # Calculate consensus sequences for alleles and output a .vcf (or .vcf.gz) with call data
   --seed 183 \  # Fixed random number generator seed for replicability
   --processes 10 \  # Number of parallel processes to use; DEFAULT: 1
-  --no-tsv  # If VCF output is enabled as above, we don't need TSV genotype output to stdout (which is the default)
 ```
 
 ##### REGARDING ALIGNMENTS
@@ -284,12 +283,12 @@ For more information, see also documentation on the [Output formats](./docs/outp
 
 If you want to output a full call report, you can use the `--json output-file.json` argument to
 specify a path to output a more detailed JSON document to. This document contains 99% CIs, peak
-labels, and some other information that isn't included in the normal TSV file. If you want this
+labels, and some other information that isn't included in the normal VCF file. If you want this
 file to be indented and human-readable, use the `--indent-json` flag in addition to `--json ...`.
 
-If you want to output a VCF file (STRs and SNVs if called; currently not phased), use the
-`--vcf ...` argument. If you pass `--vcf stdout`, the VCF will be written to `stdout` instead of a 
-file.
+VCF call information is printed to standard output by default (`--vcf stdout`). If you pass 
+`--vcf path/to/file.vcf`, the VCF will be written to a file instead. If you pass `--vcf none`, no
+VCF will be written.
 
 For more information, see also documentation on the [Output formats](./docs/output_formats.md).
 

strkit-0.25.0a3/pyproject.toml

@@ -0,0 +1,50 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "strkit"
+version = "0.25.0a3"
+dependencies = [
+    "Flask>=3.1,<3.2",
+    "orjson>=3.11.1,<4",
+    "pysam>=0.19,<0.25",
+    "numpy>=2.1.3,<3",
+    "parasail>=1.2.4,<1.4",
+    "pydantic>=2.11.7,<3",
+    "scikit-learn>=1.7.2,<2",
+    "scipy>=1.10,<1.17",
+    "statsmodels>=0.14.0,<0.15",
+    "strkit_rust_ext==0.28.0",
+]
+requires-python = ">=3.10"
+
+keywords = ["tandem repeats", "TRs", "short tandem repeats", "STRs", "genomics"]
+description = "A toolkit for analyzing variation in short(ish) tandem repeats."
+readme = "README.md"
+
+license = "GPL-3.0-only"
+license-files = ["LICEN[CS]E", "LICEN[CS]E.*"]
+classifiers = [
+    "Programming Language :: Python :: 3.10",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Programming Language :: Python :: 3.13",
+    "Programming Language :: Python :: 3.14",
+    "Programming Language :: Python :: 3 :: Only",
+    "Operating System :: POSIX",
+    "Operating System :: POSIX :: Linux",
+    "Operating System :: MacOS :: MacOS X",
+]
+
+authors = [{name = "David Lougheed", email = "david.lougheed@gmail.com"}]
+
+[project.urls]
+Homepage = "https://github.com/davidlougheed/strkit"
+Repository = "https://github.com/davidlougheed/strkit.git"
+
+[project.scripts]
+strkit = "strkit.entry:main"
+
+[tool.hatch.build]
+exclude = [".github", ".idea", "catalogs", "docs", "tests", "requirements.txt", "Dockerfile"]

strkit-0.25.0a3/strkit/__init__.py

@@ -0,0 +1,7 @@
+from importlib.metadata import version
+
+__all__ = [
+    "__version__",
+]
+
+__version__ = version("strkit")

{strkit-0.25.0a2 → strkit-0.25.0a3}/strkit/call/allele.py

@@ -18,7 +18,9 @@ from statistics import mode
 from warnings import simplefilter
 
 from numpy.typing import NDArray
-from typing import TypedDict, TYPE_CHECKING
+from typing import TYPE_CHECKING
+
+from strkit_rust_ext import CallData
 
 from .gmm import make_single_gaussian
 
@@ -31,7 +33,6 @@ if TYPE_CHECKING:
 
 __all__ = [
     "RepeatCounts",
-    "CallDict",
     "call_alleles",
 ]
 
@@ -47,11 +48,11 @@ small_allele_min = 8
 FLOAT_32_EPSILON = np.finfo(np.float32).eps
 
 
-def _calculate_cis(samples: NDArray[np.int32], is_99: bool) -> NDArray[np.int32]:
+def _calculate_cis(samples: NDArray[np.float64], is_99: bool) -> NDArray[np.int32]:
     percentiles = np.percentile(
         samples, (0.5, 99.5) if is_99 else (2.5, 97.5), axis=1, method="interpolated_inverted_cdf"
     ).transpose()
-    return percentiles
+    return np.rint(percentiles).astype(np.int32)
 
 
 def na_length_list(n_alleles: int):
@@ -126,20 +127,6 @@ def fit_gmm(
     return g
 
 
-class BaseCallDict(TypedDict):
-    call: NDArray[np.int32]
-    call_95_cis: NDArray[np.int32]
-    call_99_cis: NDArray[np.int32]
-    peaks: NDArray[np.float64]
-    peak_weights: NDArray[np.float64]
-    peak_stdevs: NDArray[np.float64]
-    modal_n_peaks: int
-
-
-class CallDict(BaseCallDict, total=False):
-    ps: int
-
-
 def get_resampled_bootstrapped_reads(
     combined_reads: NDArray[np.int32],
     combined_len: int,
@@ -204,7 +191,7 @@ def call_alleles(
     seed: int | None,
     logger_: Logger,
     debug_str: str,
-) -> CallDict | None:
+) -> CallData | None:
     combined_reads = np.concatenate((repeats_fwd, repeats_rev), axis=None, dtype=np.int32)
     combined_len = combined_reads.shape[-1]
 
@@ -217,19 +204,19 @@ def call_alleles(
 
         cn = combined_reads[0]
 
-        call = np.full(n_alleles, cn)
-        call_cis = np.full((n_alleles, 2), cn)
+        call = np.full(n_alleles, cn, dtype=np.int32)
+        call_cis = np.full((n_alleles, 2), cn, dtype=np.int32)
 
         peaks: NDArray[np.float64] = call.astype(np.float64)
 
-        return CallDict(
+        return CallData(
             call=call,
             call_95_cis=call_cis,
             call_99_cis=call_cis,
-            peaks=peaks,
-            peak_weights=np.full(n_alleles, 1.0 / n_alleles),
-            peak_stdevs=np.full(n_alleles, 0.0),
-            modal_n_peaks=1,
+            means=peaks,
+            weights=np.full(n_alleles, 1.0 / n_alleles),
+            stdevs=np.full(n_alleles, 0.0),
+            modal_n=1,
         )
 
     # ------------------------------------------------------------------------------------------------------------------
@@ -339,17 +326,17 @@ def call_alleles(
 
     peak_weights /= peak_weights.sum()  # re-normalize weights
 
-    return CallDict(
+    return CallData(
         call=medians_of_means_final.flatten(),
         call_95_cis=allele_cis_95,
         call_99_cis=allele_cis_99,
 
-        peaks=medians_of_means.flatten(),  # Don't round, so we can recover original Gaussian model
-        peak_weights=peak_weights,
-        peak_stdevs=peak_stdevs.flatten(),
+        means=medians_of_means.flatten(),  # Don't round, so we can recover original Gaussian model
+        weights=peak_weights,
+        stdevs=peak_stdevs.flatten(),
         # TODO: should be ok to use this, because resample gets put at end, vertically (3rd allele in a 3-ploid case)
         #  so taking the first 2 alleles still works in terms of stdev/mean estimates? I think?
         #  Not quite, cause it's sorted...
         #  --> Only do the peak assignment with 1/2 peaks, which is the majority of human situations
-        modal_n_peaks=modal_n_peaks,
+        modal_n=modal_n_peaks,
     )