BBEst

Basic Usage

bbest.sh in=<sam file> out=<stats file>

This tool processes SAM alignment files containing mapped ESTs (Expressed Sequence Tags) to generate mapping statistics and capture metrics. ESTs are often broken into smaller pieces during sequencing and mapping; this tool reassembles them and evaluates assembly capture efficiency.

Parameters

Parameters control input/output files, reference sequences, and mapping quality thresholds for EST capture analysis.

Input/Output Parameters

in=<file>: Specify a sam file (or stdin) containing mapped ESTs. The SAM file should be generated with BBMap using specific flags: k=13 maxindel=100000 customtag ordered. The 'ordered' flag is particularly important as the tool expects ESTs to appear in input order to properly reassemble multi-part sequences.
out=<file>: Specify the output stats file (default is stdout). The output contains detailed statistics about EST mapping including counts and percentages for different mapping categories, intron analysis, and capture efficiency metrics.
ref=<file>: Specify the reference file (optional). When provided, allows additional validation and analysis of the assembly used as mapping target.
est=<file>: Specify the EST fasta file (optional). When provided, enables additional analysis comparing original EST sequences with mapping results.

Analysis Parameters

fraction=0.98: Minimum fraction of bases mapped to reference to be considered 'all mapped'. ESTs with mapping coverage at or above this threshold are categorized as having all bases successfully captured by the assembly. Default value of 0.98 means 98% of EST bases must map to qualify as fully captured.

Examples

Basic EST Analysis

bbest.sh in=mapped_ests.sam out=est_stats.txt

Analyze EST mapping from a SAM file and output capture statistics to a text file.

Complete EST Analysis with Reference

bbest.sh in=mapped_ests.sam out=est_stats.txt ref=assembly.fasta est=original_ests.fasta

Analysis including reference assembly and original EST sequences for validation.

Custom Mapping Threshold

bbest.sh in=mapped_ests.sam out=est_stats.txt fraction=0.95

Use a lower threshold (95%) for considering ESTs as fully captured, which may be appropriate for more fragmented assemblies.

Processing from Standard Input

samtools view alignment.bam | bbest.sh in=stdin out=est_capture.txt

Process SAM data directly from a pipeline, useful for integrating with other tools.

Algorithm Details

EST Reassembly Strategy

The tool implements pattern-based EST reassembly using string parsing and HashMap tracking that handles the common scenario where long EST sequences are broken into smaller pieces for mapping. The algorithm:

Part Detection: Identifies EST parts using the naming convention "_part[number]" suffix, extracting the base EST name
Sequential Assembly: Requires input SAM file to be ordered (using BBMap 'ordered' flag) to properly group parts belonging to the same EST
Length Calculation: Accumulates total length from all parts to reconstruct original EST size
Mapping Assessment: Evaluates mapping quality across all parts to determine overall EST capture success

Mapping Quality Categories

ESTs are classified into four mapping quality categories based on the fraction of bases successfully mapped:

All Bases Mapped: ESTs with ≥fraction (default 98%) of bases mapped to the reference
Most Bases Mapped: ESTs with ≥50% but <fraction of bases mapped
Some Bases Mapped: ESTs with >0% but <50% of bases mapped
Zero Bases Mapped: ESTs with no successful mappings

Multi-Scaffold Analysis

The tool tracks ESTs that map to multiple scaffolds, which can indicate:

Assembly fragmentation where single genes are split across scaffolds
Chimeric EST sequences
Repetitive sequences present in multiple locations

Intron Detection and Analysis

Splice junction analysis is performed by parsing CIGAR strings for deletion (D) and skipped region (N) operations:

Intron Size Distribution: Calculates minimum, maximum, median, and average intron sizes
Splice Site Validation: Large deletions/skips likely represent genuine introns in eukaryotic sequences
Threshold Filtering: Applies minimum intron size filter (default 10bp) to reduce noise from small indels

Output Statistics Format

The tool generates output statistics including:

Count Metrics: Number of reference scaffolds, ESTs processed, and base counts
Percentage Metrics: Capture efficiency percentages for each mapping category
Intron Statistics: Complete intron size distribution analysis
Multi-scaffold Metrics: Count and percentage of ESTs mapping to multiple scaffolds

Memory Efficiency

The algorithm uses specific data structures to manage memory usage:

Streaming Processing: Processes SAM file line-by-line without loading entire file into memory
Hash-based Tracking: Uses HashMap<String, EST> for EST lookup and grouping by name
Minimal Memory Footprint: Default memory allocation of only 120MB for most analyses

Performance Characteristics

The tool has the following processing characteristics:

Linear Time Complexity: O(n) processing time where n is the number of SAM records
Scalable Design: Processes records sequentially with constant memory overhead per EST
I/O Implementation: Uses TextFile and TextStreamWriter classes for SAM input and statistics output

Output Format

The output statistics file contains the following sections:

File Information

ref_file: Path to reference file (if provided)
est_file: Path to EST file (if provided)
sam_file: Path to input SAM file

Count Statistics

n_ref_scaffolds: Number of reference scaffolds
n_ref_bases: Total reference bases
n_est: Total number of ESTs processed
n_est_bases: Total EST bases

Mapping Categories

Each category shows: type, n_est, pct_est, n_bases, pct_bases

all: ESTs with ≥fraction bases mapped
most: ESTs with ≥50% but <fraction bases mapped
some: ESTs with >0% but <50% bases mapped
zero: ESTs with no mapped bases
multi: ESTs mapping to multiple scaffolds

Intron Analysis

Final line contains: count, min, max, median, average intron sizes

Preprocessing Requirements

For optimal results, SAM files should be generated with specific BBMap parameters:

Required BBMap Flags

k=13: Short kmer size appropriate for EST mapping
maxindel=100000: Allow large indels to capture introns and structural variations
customtag: Add custom tags for detailed alignment information
ordered: Maintain input order crucial for EST part reassembly

EST Naming Convention

For proper part reassembly, EST parts should follow the naming pattern:

[EST_NAME]_part[NUMBER]

Example: EST12345_part1, EST12345_part2, EST12345_part3

Support

For questions and support:

Email: bbushnell@lbl.gov
Documentation: bbmap.org