Biopython Documentation

A SeqRecord object is the primary container in Biopython used to hold a biological sequence (as a Seq object) along with its associated metadata. It includes identifiers (id and name), a description, and optional annotations and features (sub-features).

While most users obtain SeqRecord objects by parsing files using the Bio.SeqIO module, they can also be instantiated manually.

Key components of a SeqRecord include:

• Sequence: The actual biological sequence, stored as a Bio.Seq.Seq object.
• Identifiers: String attributes for id, name, and description.
• Annotations: A dictionary containing metadata (e.g., taxonomy, source, or references).
• Features: A list of SeqFeature objects representing specific regions of the sequence.
• Database Cross-references: A list of database identifiers (dbxrefs).

from Bio.Seq import Seq
from Bio.SeqRecord import SeqRecord

record = SeqRecord(
    Seq("MKQHKAMIVALIVICITAVVAALVTRKDLCEVHIRTGQTEVAVF"),
    id="YP_025292.1",
    name="HokC",
    description="toxic membrane protein, small",
)
print(record)

When implementing new statistics for the Population Genetics module, developers must account for several distinct dimensions to ensure the framework is flexible enough for different research communities (e.g., cancer studies vs. conservation).

Key dimensions to consider include:

• Marker Dependency: Determine if the statistic is marker independent (e.g., $F_{ST}$ which relies on allele counts) or marker dependent (e.g., Tajima's D which requires sequences/RFLPs, or Allelic range which requires microsatellites/STRs).
• Population Scope: Distinguish between Intra-Population statistics (e.g., observed heterozygosity, which has no notion of population structure) and Inter-population statistics (e.g., $F_{ST}$, used to measure population differentiation).
• Data Type: Identify the required input format:
  • Haplotypic
  • Genotypic (phase known)
  • Genotypic (phase unknown)
  • Genotypic dominant
  • Frequency only (e.g., sufficient for expected heterozygosity, but not for observed heterozygosity).
• Locus Scale: Specify if the statistic is Single locus (e.g., allelic richness, $F_{ST}$) or Multi-loci (e.g., linkage disequilibrium (LD) or EHH).
• Temporal Aspect: Determine if the statistic is Temporal/longitudinal (e.g., temporal-$F_{ST}$) or a Single point in time.

A robust implementation should support a wide range of marker densities and dataset sizes, from a few loci to millions of SNPs.

Bio.SearchIO uses standardized conventions for coordinates to ensure consistency across different search formats and to allow direct use with Python slicing.

Python-style coordinates

Coordinates use zero-based, half-open intervals.

• Example: If a BLAST XML file reports HSP start/end as 10 and 28, Bio.SearchIO stores them as 9 and 28.
• This allows you to slice a sequence directly: sequence[start:end].
• When writing to a file via Bio.SearchIO.write, these values are automatically converted back to the original format's convention.

Coordinate order

In some formats (like BLAST plain text on the minus strand), the start coordinate is larger than the end coordinate.

• In Bio.SearchIO, start is always smaller than end, regardless of the strand.
• This convention is enforced at the HSPFragment level. The order of fragments within an HSP object still follows the original file's order.
• Like Python-style coordinates, these are restored to the original format's convention upon writing.

The _BitString class is an internal utility in Bio.Phylo.Consensus used to represent clades as binary strings. This allows for efficient counting and comparison of clades across multiple trees.

Capabilities:

• Binary Operations: Supports & (AND), | (OR), ^ (XOR), and ~ (NOT).
• Inspection: index_one() returns indices of '1's; index_zero() returns indices of '0's.
• Comparison: contains(), independent(), and iscompatible() methods for bitstring relationship testing.

Manual Clade to BitString Conversion: To represent a clade as a _BitString relative to a set of terminal names:

1. Get all terminal names from the reference tree.
2. For a specific clade, check which terminal names it contains.
3. Construct a string of '1's and '0's based on those matches.

# Example: Creating a bitstring for a clade
term_names = [term.name for term in trees[0].get_terminals()]
clade_term_names = [term.name for term in clade.get_terminals()]
boolvals = [name in clade_term_names for name in term_names]
bitstr = _BitString("".join(map(str, map(int, boolvals))))