Support to read genomic ranges objects

setup.cfg

@@ -55,6 +55,7 @@ install_requires =
     scipy
     singlecellexperiment>=0.4.1
     summarizedexperiment>=0.4.1
+    genomicranges>=0.4.9
 
 [options.packages.find]
 where = src

src/rds2py/granges.py

@@ -0,0 +1,111 @@
+from genomicranges import GenomicRanges, SeqInfo
+from iranges import IRanges
+from biocframe import BiocFrame
+
+from .parser import get_class
+from .pdf import as_pandas_from_dframe
+
+__author__ = "jkanche"
+__copyright__ = "jkanche"
+__license__ = "MIT"
+
+
+def as_granges(robj):
+    """Parse an R object as a :py:class:`~genomicranges.GenomicRanges.GenomicRanges`.
+
+    Args:
+        robj:
+            Object parsed from the `RDS` file.
+
+            Usually the result of :py:func:`~rds2py.parser.read_rds`.
+
+    Returns:
+        A ``GenomicRanges`` object.
+    """
+    _cls = get_class(robj)
+
+    if _cls not in ["GenomicRanges", "GRanges"]:
+        raise TypeError(f"obj is not genomic ranges, but is `{_cls}`.")
+
+    _range_start = robj["attributes"]["ranges"]["attributes"]["start"]["data"]
+    _range_width = robj["attributes"]["ranges"]["attributes"]["width"]["data"]
+    _range_names = None
+    if "NAMES" in robj["attributes"]["ranges"]["attributes"]:
+        _range_names = robj["attributes"]["ranges"]["attributes"]["NAMES"]["data"]
+    _ranges = IRanges(_range_start, _range_width, names=_range_names)
+
+    _seqnames = _as_list(robj["attributes"]["seqnames"])
+
+    _strand_obj = robj["attributes"]["strand"]["attributes"]["values"]
+    _strands = _strand_obj["data"]
+    if "attributes" in _strands:
+        if "levels" in _strands["attributes"]:
+            _levels_data = _strands["attributes"]["levels"]["data"]
+            _strands = [_levels_data[x] for x in _strands]
+
+    _seqinfo_seqnames = robj["attributes"]["seqinfo"]["attributes"]["seqnames"]["data"]
+    _seqinfo_seqlengths = robj["attributes"]["seqinfo"]["attributes"]["seqlengths"][
+        "data"
+    ]
+    _seqinfo_is_circular = robj["attributes"]["seqinfo"]["attributes"]["is_circular"][
+        "data"
+    ]
+    _seqinfo_genome = robj["attributes"]["seqinfo"]["attributes"]["genome"]["data"]
+    _seqinfo = SeqInfo(
+        seqnames=_seqinfo_seqnames,
+        seqlengths=[None if x == -2147483648 else x for x in _seqinfo_seqlengths],
+        is_circular=[None if x == -2147483648 else x for x in _seqinfo_is_circular],
+        genome=_seqinfo_genome,
+    )
+
+    _mcols = BiocFrame.from_pandas(
+        as_pandas_from_dframe(robj["attributes"]["elementMetadata"])
+    )
+
+    _gr_names = None
+    if "NAMES" in robj["attributes"]:
+        _gr_names = robj["attributes"]["NAMES"]["data"]
+
+    return GenomicRanges(
+        seqnames=_seqnames,
+        ranges=_ranges,
+        names=_gr_names,
+        mcols=_mcols,
+        seqinfo=_seqinfo,
+    )
+
+
+def _as_list(robj):
+    """Parse an R object as a :py:class:`~list`.
+
+    Args:
+        robj:
+            Object parsed from the `RDS` file.
+
+            Usually the result of :py:func:`~rds2py.parser.read_rds`.
+
+    Returns:
+        A ``list`` of the Rle class.
+    """
+    _cls = get_class(robj)
+
+    if _cls not in ["Rle"]:
+        raise TypeError(f"obj is not Rle, but is `{_cls}`.")
+
+    _attr_vals = robj["attributes"]
+    _data = _attr_vals["values"]["data"].tolist()
+    if "attributes" in _attr_vals["values"]:
+        if "levels" in _attr_vals["values"]["attributes"]:
+            _levels_data = _attr_vals["values"]["attributes"]["levels"]["data"]
+            _data = [_levels_data[x - 1] for x in _data]
+
+    if "lengths" in _attr_vals:
+        _final = []
+        _lengths = _attr_vals["lengths"]["data"]
+
+        for idx, lg in enumerate(_lengths.tolist()):
+            _final.extend([_data[idx]] * lg)
+
+        _data = _final
+
+    return _data

tests/data/generate_files.R

@@ -95,3 +95,14 @@ saveRDS(y, file="s4_matrix.rds")
 setClass("FOO", slots=c(bar="integer"))
 y <- new("FOO", bar=2L)
 saveRDS(y, file="s4_class.rds")
+
+# GenomicRanges
+
+gr <- GRanges(
+    seqnames = Rle(c("chr1", "chr2", "chr1", "chr3"), c(1, 3, 2, 4)),
+    ranges = IRanges(101:110, end = 111:120, names = head(letters, 10)),
+    strand = Rle(strand(c("-", "+", "*", "+", "-")), c(1, 2, 2, 3, 2)),
+    score = 1:10,
+    GC = seq(1, 0, length=10))
+
+saveRDS(gr, file="granges.rds")

tests/test_granges.py

@@ -0,0 +1,18 @@
+import pytest
+
+from rds2py.granges import as_granges
+from rds2py.parser import read_rds
+
+from genomicranges import GenomicRanges
+
+__author__ = "jkanche"
+__copyright__ = "jkanche"
+__license__ = "MIT"
+
+
+def test_granges():
+    robj = read_rds("tests/data/granges.rds")
+
+    gr = as_granges(robj=robj)
+
+    assert isinstance(gr, GenomicRanges)