Sumstats — Filter

Sumstats object containing the data to filter.

Query expression using pandas.DataFrame.query syntax

If True, removes variants meeting the condition

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table. When called via :meth:Sumstats.filter_value(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object containing the data to filter.

Dictionary of {column: threshold} for lower bounds (variant values < threshold will be kept)

Dictionary of {column: threshold} for upper bounds (variant values > threshold will be kept)

Dictionary of {column: value} for equality checks (variant values == value will be kept)

If True, removes variants meeting the condition

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table. When called via :meth:Sumstats.filter_in(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object containing the data to filter.

Dictionary of {column: threshold} for lower bounds (variant values < threshold will be removed)

Dictionary of {column: threshold} for upper bounds (variant values > threshold will be removed)

Dictionary of {column: value} for equality checks (variant values == value will be removed)

If True, removes variants meeting the condition

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table

Sumstats object or DataFrame to process.

Genomic region to filter. Can be: - List/tuple [chr, start, end]: Specific genomic region coordinates - String formats (normalized automatically): - "chr:start-end" (e.g., "chr1:1000-5000") - "chr:pos:flanking" (e.g., "1:5000:2000" or "1:5000:2kb") - "snpid:flanking" (e.g., "rs123:500" or "1:5000:C:T:2000") - Special strings: - "HLA": Filter variants in HLA region (excludes HLA) - "HIGH_LD": Filter variants in high LD regions (excludes high LD) - "PAR": Filter variants in pseudo-autosomal regions (excludes PAR)

Column name for chromosome information

Column name for position information

Genome build version (19 or 38) for HLA/HIGH_LD/PAR regions

If True, writes progress to log

Logger instance

Subset of summary statistics in the specified region (or with specified regions excluded). When called via :meth:Sumstats.filter_region(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object or DataFrame to process.

Path to BED file containing regions of interest

Column name for chromosome information

Column name for position information

If True, uses high LD regions from the specified build

Genome build version (19 or 38) for high LD regions

If True, writes progress to log

Logger instance

Filtered summary statistics table containing only variants in the specified regions. When called via :meth:Sumstats.filter_region_in(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object or DataFrame to process.

Path to BED file containing regions to exclude

Column name for chromosome information

Column name for position information

If True, excludes variants in high LD regions from the specified build

Genome build version (19 or 38) for high LD regions

If True, writes progress to log

Logger instance

Filtered summary statistics table with variants in specified regions removed. When called via :meth:Sumstats.filter_region_out(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object or DataFrame to process.

ID of the central variant (must exist in SNPID column)

Size of flanking region in kilobases

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Variants in flanking regions. When called via :meth:Sumstats.filter_flanking(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Chromosome and position coordinate(s) to use as center

Size of flanking region in kilobases

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Variants in flanking regions. When called via :meth:Sumstats.filter_flanking_by_chrpos(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Sumstats object or DataFrame to process.

Variant ID(s) to use as center (searches in rsID/SNPID columns)

Size of flanking region in kilobases

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Variants in flanking regions. When called via :meth:Sumstats.filter_flanking_by_id(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

"in" to keep palindromic variants, "out" to remove them

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table. When called via :meth:Sumstats.filter_palindromic(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

"in" to keep SNPs, "out" to remove SNPs

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table. When called via :meth:Sumstats.filter_snp(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

"in" to keep indels, "out" to remove indels

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table. When called via :meth:Sumstats.filter_indel(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Input summary statistics.

rsID column name.

Chromosome column name.

Base-pair position column name.

Effect allele column name.

Non-effect allele column name.

Genome build ("19" or "38").

Print progress messages.

Require allele agreement with the HapMap3 reference.

pandas merge type (inner, left, etc.).

Logging object.

Rows matching HapMap3 variants.

Column name for chromosome information

Column name for position information

Lower bound of genomic region

Upper bound of genomic region

Genome build version (19 or 38)

"xmhc" for extended MHC region, "hla" or "mhc" for classical HLA region

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None.

Filtered summary statistics table with HLA variants removed. When called via :meth:Sumstats.exclude_hla(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

GWAS summary statistics table

List of variant identifiers to search for. Accepts multiple formats: - CHR:POS (e.g., '1:123456') - CHR-POS (e.g., '1_123456') - rsID (e.g., 'rs12345') - SNPID (e.g., '1:123456:A:G') - List of [CHR, POS] - Full variant IDs with alleles (e.g., 'chr1:123456:A:G')

If True, writes progress to log

Subset of summary statistics containing matching variants. When called via :meth:Sumstats.search(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

Number of variants to sample

Fraction of variants to sample (alternative to n)

If True, writes progress to log

If False, return a new Sumstats object containing the filtered results. If True, apply the filter to the current object in place and return None. **kwargs : dict Additional arguments for pandas.DataFrame.sample

Subsampled summary statistics table. When called via :meth:Sumstats.random_variants(), returns a new Sumstats object if inplace=False, or updates the Sumstats object in place (modifies self.data) and returns None if inplace=True.

List of full SNPIDs to find proxies for.

Sumstats object or DataFrame with summary statistics for existing SNPs.

Path to the VCF file for reference genotypes.

If True, write detailed logs.

Size in kb for the flanking region around each SNP.

Minimum R^2 value to consider a proxy valid.

If True, include proxy variants from VCF that are not in common_sumstats. When False, only returns proxies that are already in common_sumstats.

Extracted summary statistics including matched proxies sorted by LD strength.

Chromosome identifier.

Base-pair position.

Window size in kilobases.

Print log message.

[chrom, start, end], where start and end are base-pair coordinates (int).