echtvar-hail

echtvar's variant encodings in Hail, and an export that writes them to Parquet instead of echtvar's zip container.

Two things live here:

• var32 / kmer16 — Python ports of src/lib/{var32,kmer16}.rs, each with an hl_* twin so the encodings run inside a Hail pipeline rather than by collecting rows to Python.
• parquet — an export in echtvar's per-(chrom, chunk) shape, as standard Parquet. The archive becomes a self-describing columnar file that DuckDB, Polars, pandas or Arrow can read without echtvar.

Install

pip install echtvar-hail

Python 3.11 or 3.12. Hail brings PySpark, which needs a JVM — Hail is built and tested against Java 11.

Quick start

One Hail table, one field config, covering every shape Field supports — copied as-is, quantized, renamed, and a nested struct — plus the routing that splits variants across the two output tables.

import hail as hl
import pyarrow.parquet as pq
from echtvar_hail import var32
from echtvar_hail.parquet import Field, export_parquet, finalize_parquet

hl.init()

ht = hl.Table.parallelize(
    [
        {"locus": hl.Locus("chr1", 100, "GRCh38"), "alleles": ["A", "T"],
         "filters": "PASS", "af": 0.0031, "cadd_raw": 3.21,
         "vep": [hl.Struct(consequence="missense_variant", biotype="protein_coding", polyphen=0.91),
                 hl.Struct(consequence="intron_variant", biotype="lncRNA", polyphen=0.05)]},
        {"locus": hl.Locus("chr1", 200, "GRCh38"), "alleles": ["AC", "A"],
         "filters": "PASS", "af": 0.41, "cadd_raw": -1.44,
         "vep": [hl.Struct(consequence="intron_variant", biotype="protein_coding", polyphen=0.01)]},
        {"locus": hl.Locus("chr1", 300, "GRCh38"), "alleles": ["A", "ACGTACGT"],
         "filters": "low_qual", "af": 0.0001, "cadd_raw": 22.6,
         "vep": [hl.Struct(consequence="stop_gained", biotype="protein_coding", polyphen=1.0)]},
    ],
    hl.tstruct(
        locus=hl.tlocus("GRCh38"), alleles=hl.tarray(hl.tstr),
        filters=hl.tstr, af=hl.tfloat64, cadd_raw=hl.tfloat64,
        vep=hl.tarray(hl.tstruct(consequence=hl.tstr, biotype=hl.tstr, polyphen=hl.tfloat64)),
    ),
)

fields = [
    Field("filters"),                                  # plain: copied as-is
    Field("af", precision=1e-6),                        # quantized: to a stated accuracy
    Field("cadd_raw", alias="cadd", precision=0.01),    # quantized + renamed
    Field("vep", fields=(                               # struct: member-by-member
        Field("consequence", categorical=True),
        Field("biotype", categorical=True),
        Field("polyphen", precision=1e-3),
    )),
]

Every Hail column not named in fields is dropped — locus/alleles become the key, and anything else absent from this list simply doesn't reach the file. What each Field does to its column:

# Stage 1 (Hail + Spark, distributed): route short/long, encode, partition.
short, long = export_parquet(ht, "/tmp/staging", fields=fields)

# Stage 2 (pyarrow, one node): choose encodings, consolidate.
finalize_parquet(short, "/tmp/echtvar/short", fields=fields)
finalize_parquet(long, "/tmp/echtvar/long", fields=fields)

Reading it back — a key plus its chrom/chunk is a whole variant, quantized values divide back out by their multiplier, and the struct comes back as a list of dicts:

t = pq.read_table("/tmp/echtvar/short/chr1.parquet")
print(t.schema.names)
# ['key', 'filters', 'af', 'cadd', 'vep', 'chrom', 'chunk']

for i in range(t.num_rows):
    d = var32.decode(t["key"][i].as_py(), t["chrom"][i].as_py(), t["chunk"][i].as_py())
    print(f"{d.chrom}:{d.position + 1} {d.reference}>{d.alternate}")
    print(f"  filters={t['filters'][i].as_py()!r}  af={t['af'][i].as_py() / 1_000_000}"
          f"  cadd={t['cadd'][i].as_py() / 100}")
    print(f"  vep={t['vep'][i].as_py()}")

# chr1:100 A>T
#   filters='PASS'  af=0.0031  cadd=3.21
#   vep=[{'consequence': 'missense_variant', 'biotype': 'protein_coding', 'polyphen': 910},
#        {'consequence': 'intron_variant',   'biotype': 'lncRNA',         'polyphen': 50}]
# chr1:200 AC>A
#   filters='PASS'  af=0.41  cadd=-1.44
#   vep=[{'consequence': 'intron_variant', 'biotype': 'protein_coding', 'polyphen': 10}]

The third variant is missing from that output because A>ACGTACGT exceeds var32's four-base budget, so it routed to the long table instead:

t = pq.read_table("/tmp/echtvar/long/chr1.parquet")
print(t.schema.field("key").type, t.num_rows)
# list<element: uint32> 1

Why two stages

No single tool can write this file. Hail and Spark are needed for the encoding and the shuffle, at scale. But Spark writes Parquet through parquet-mr, which cannot be told a column's encoding at all — it infers one from a writer version and a dictionary setting. The choices that make the format worth having are unavailable there.

Stage 2 output is one file per chromosome, several row groups per chunk, so a chunk lookup is a row-group read rather than opening a file. Measured over 400 chunks: same bytes to within 1%, and 2.9× faster on random chunk reads locally — the gap widens on object storage, where opening a file is a round trip.

API

Field

How one annotation is carried. Three shapes, and any of them may be array-valued — VEP-style annotations usually are, and an array stays an array rather than being flattened or joined into a delimited string.

Field("af")                              # carried as-is
Field("af", precision=1e-6)              # float -> int, to a stated accuracy
Field("consequence", categorical=True)   # dictionary-encoded (echtvar's string table)
Field("cadd_raw", alias="cadd", precision=0.01)   # renamed on the way out

Nested fields

The vep field above is the common case — nest fields to describe a struct's members, an array of them being one per transcript. Here's why it's shaped that way:

Field("csq", fields=(
    Field("consequence", categorical=True),
    Field("biotype", alias="tx_biotype", categorical=True),
    Field("polyphen", precision=1e-3),
))

Given [hl.struct(consequence="missense_variant", biotype="protein_coding", polyphen=0.912), ...] that writes one Parquet leaf per member:

csq.list.element.consequence   BYTE_ARRAY  RLE_DICTIONARY
csq.list.element.tx_biotype    BYTE_ARRAY  RLE_DICTIONARY
csq.list.element.polyphen      INT64       PLAIN            # 0.912 -> 912

Three things follow from each member being its own column:

• Each takes its own encoding. The two string members get their own dictionaries; the quantized one correctly gets none.

• fields is an allowlist. Members you don't name are dropped, which is how you take three columns out of a VEP annotation with forty.

• precision applies per member, and the export's sentinel check reaches inside:

  export_parquet(ht, staging, fields=fields)
  # [{'column': 'csq.polyphen', 'multiplier': 1000, 'min': 0.05, 'max': 1.0, ...}]
  

The struct survives to disk as a struct — nothing is flattened or exploded, so a row reads back as a list of dicts:

t["csq"][0].as_py()
# [{'consequence': 'missense_variant', 'tx_biotype': 'protein_coding', 'polyphen': 912},
#  {'consequence': 'intron_variant',   'tx_biotype': 'lncRNA',         'polyphen': 50}]

The array is incidental. A plain struct nests the same way, to any depth — fields describes the shape, and whether it sits inside a list only changes the leaf paths:

Field("a", fields=(
    Field("b", fields=(Field("c", categorical=True), Field("score", precision=1e-3))),
    Field("n"),
))

a.b.c       BYTE_ARRAY  RLE_DICTIONARY
a.b.score   INT64       PLAIN            # 0.125 -> 125
a.n         INT32       PLAIN

# arrow type: struct<b: struct<c: string, score: int64>, n: int32>
# the nested leaf a.b.score is quantized with multiplier 1000

precision is an accuracy budget, not a performance knob. Quantization shrinks a column by making values repeat, so what you keep is what you don't save. On 300k gnomAD-like allele frequencies, against 1.09MB as float32:

An allele frequency is AC/AN, so it cannot mean anything below 1/AN — about 6.7e-7 at AN≈1.5M. Asking for finer spends bytes on the denominator wobbling between sites.

export_parquet(ht, base_path, *, locus="locus", alleles="alleles", fields=(), mode="overwrite")

Stage 1. Returns (short_path, long_path). Before writing, it refuses any precision whose observed range would reach the missing sentinel — that failure is otherwise silent, since real values would simply read back as absent. One aggregation pass, and free when no field is quantized. Routes each variant on len(ref) + len(alt) > 4: short variants take a 32-bit var32 key, long ones a kmer16 array. Multi-allelic rows contribute only their first alternate — split them (hl.split_multi_hts) first.

finalize_parquet(staging_path, output_path, *, fields=(), filesystem=None)

Stage 2. Pass the same fields. On GCS, hand it a pyarrow.fs.GcsFileSystem — pyarrow does not use Hadoop's GCS connector.

Decoding

var32.decode(key, chrom, chunk) and kmer16.decode(key, chrom) return a PRA(chrom, chunk, position, reference, alternate), 0-based. A var32 key stores position truncated to 20 bits, so it needs its chunk to reconstruct an absolute coordinate — which is why chunk is a column.

Output layout

out/short/chr1.parquet      key: uint32          16k-row row groups, nested in chunks
out/long/chr1.parquet       key: list<uint32>

Columns are key, chrom, chunk, and one per field. Position and alleles are not stored — the key plus chunk reconstructs them.

Row groups hold ~16,000 rows and nest inside chunks — a chunk becomes several of them, and none ever spans a chunk boundary. A lookup is two stages, and the order matters:

f = pq.ParquetFile("out/short/chr1.parquet")
md = f.metadata
ci, ki = md.schema.names.index("chunk"), md.schema.names.index("key")
want, key = chunk_of(pos0), var32.encode(pos0, ref, alt)

def holds(i):
    rg = md.row_group(i)
    k = rg.column(ki).statistics
    return rg.column(ci).statistics.min == want and k.min <= key <= k.max

rg = next(i for i in range(md.num_row_groups) if holds(i))   # footer only, no I/O
keys = f.read_row_group(rg, columns=["key"])["key"].to_pylist()
i = bisect.bisect_left(keys, key)

Stage 1 cannot be skipped. var32 keys overlap across chunks by construction (position is truncated to POSITION_BITS), so matching on key statistics alone lands in another chunk's row group. Within a chunk the rows stay key-ordered, so its row groups have disjoint ascending key ranges and exactly one can hold the key. And never index positionally — a chunk with no variants produces no row group at all.

Row-group size is what sets lookup cost, because pyarrow decodes a whole row group either way (it prunes on the page index for neither read_row_group() nor a filtered scan). Measured on 2.5M real gnomAD chr1 rows:

16,000 sits before the knee: 25× faster lookups for 2% more space. Below it compression collapses — encodings and zstd's window reset per row group — and footer parsing starts to dominate.

chrom identifies rows once several files are read together (ds.dataset(root) concatenates them and does not surface filenames). Both are constant over their scope, so RLE reduces them to almost nothing.

Quantized columns are unreadable without their multiplier, so it is written to the Parquet footer under the fields key:

import json
md = pq.ParquetFile("/tmp/echtvar/short/chr1.parquet").metadata
json.loads(md.metadata[b"fields"])
# [{'name': 'filters', 'column': 'filters'},
#  {'name': 'af', 'column': 'af', 'multiplier': 1000000, 'missing': 4294967295},
#  {'name': 'cadd_raw', 'column': 'cadd', 'multiplier': 100, 'missing': 4294967295},
#  {'name': 'vep', 'column': 'vep', 'fields': [
#      {'name': 'consequence', 'column': 'consequence'},
#      {'name': 'biotype', 'column': 'biotype'},
#      {'name': 'polyphen', 'column': 'polyphen', 'multiplier': 1000, 'missing': 4294967295}]}]

A struct field's own entry has no multiplier — it's a container, so the multiplier sits on the member that needs one, and the nesting in the config is mirrored in the footer.

Dictionary encoding is deliberately not recorded there — Parquet already reports each column chunk's encoding, and a footer note would be the less reliable of the two.

Development

uv sync            # dev tools are a PEP 735 dependency group
uv run pytest
uv run ruff check

The suite starts a real Hail/Spark session; it needs a JVM.