Version Compatibility

Reference examples tested with: BioPython 1.83+

Before using code patterns, verify installed versions match. If versions differ:

• Python: pip show <package> then help(module.function) to check signatures

If code throws ImportError, AttributeError, or TypeError, introspect the installed package and adapt the example to match the actual API rather than retrying.

FASTQ Quality Scores

"Filter my FASTQ reads by quality score" → Access, analyze, and filter Phred quality scores, trim low-quality bases, and generate per-position quality profiles.

• Python: SeqIO.parse() with letter_annotations['phred_quality'] (BioPython)

Analyze and manipulate FASTQ quality scores using Biopython.

Required Imports

python
from Bio import SeqIO
from Bio.Seq import Seq

Accessing Quality Scores

Quality scores are stored in letter_annotations['phred_quality'] as a list of integers.

python
for record in SeqIO.parse('reads.fastq', 'fastq'):
    qualities = record.letter_annotations['phred_quality']
    print(record.id, qualities[:10])  # First 10 quality scores

Quality Score Basics

| Phred Score | Error Probability | Accuracy | |-------------|-------------------|----------| | 10 | 1 in 10 | 90% | | 20 | 1 in 100 | 99% | | 30 | 1 in 1000 | 99.9% | | 40 | 1 in 10000 | 99.99% |

Code Patterns

Calculate Average Quality per Read

python
for record in SeqIO.parse('reads.fastq', 'fastq'):
    quals = record.letter_annotations['phred_quality']
    avg_qual = sum(quals) / len(quals)
    print(f'{record.id}: {avg_qual:.1f}')

Filter Reads by Mean Quality

python
def high_quality_reads(records, min_avg_qual=20):
    for record in records:
        quals = record.letter_annotations['phred_quality']
        if sum(quals) / len(quals) >= min_avg_qual:
            yield record

records = SeqIO.parse('reads.fastq', 'fastq')
good_reads = high_quality_reads(records, min_avg_qual=25)
SeqIO.write(good_reads, 'filtered.fastq', 'fastq')

Filter by Minimum Quality at Any Position

python
def all_bases_above(records, min_qual=20):
    for record in records:
        if min(record.letter_annotations['phred_quality']) >= min_qual:
            yield record

Trim Low-Quality Ends (3' Trimming)

python
def trim_low_quality(record, min_qual=20):
    quals = record.letter_annotations['phred_quality']
    trim_pos = len(quals)
    for i in range(len(quals) - 1, -1, -1):
        if quals[i] >= min_qual:
            trim_pos = i + 1
            break
    return record[:trim_pos]

records = SeqIO.parse('reads.fastq', 'fastq')
trimmed = (trim_low_quality(rec) for rec in records)
SeqIO.write(trimmed, 'trimmed.fastq', 'fastq')

Sliding Window Quality Trim

Goal: Trim a read at the first position where average quality in a sliding window drops below threshold.

Approach: Slide a fixed-size window across quality scores; when the window average falls below the cutoff, truncate the record at that position.

Reference (BioPython 1.83+):

python
def sliding_window_trim(record, window_size=5, min_avg_qual=20):
    quals = record.letter_annotations['phred_quality']
    for i in range(len(quals) - window_size + 1):
        window = quals[i:i + window_size]
        if sum(window) / window_size < min_avg_qual:
            return record[:i] if i > 0 else None
    return record

Quality Statistics Summary

python
import statistics

all_quals = []
for record in SeqIO.parse('reads.fastq', 'fastq'):
    all_quals.extend(record.letter_annotations['phred_quality'])

print(f'Mean quality: {statistics.mean(all_quals):.1f}')
print(f'Median quality: {statistics.median(all_quals):.1f}')
print(f'Min quality: {min(all_quals)}')
print(f'Max quality: {max(all_quals)}')

Per-Position Quality Profile

Goal: Compute mean quality at each read position to identify systematic quality drops (e.g., read-end degradation).

Approach: Accumulate quality scores by position across all reads, then compute per-position means.

Reference (BioPython 1.83+):

python
from collections import defaultdict

position_quals = defaultdict(list)
for record in SeqIO.parse('reads.fastq', 'fastq'):
    for i, q in enumerate(record.letter_annotations['phred_quality']):
        position_quals[i].append(q)

for pos in sorted(position_quals.keys())[:20]:
    quals = position_quals[pos]
    print(f'Position {pos}: mean={sum(quals)/len(quals):.1f}')

Count Reads by Quality Threshold

python
thresholds = [20, 25, 30, 35]
counts = {t: 0 for t in thresholds}

for record in SeqIO.parse('reads.fastq', 'fastq'):
    avg = sum(record.letter_annotations['phred_quality']) / len(record.seq)
    for t in thresholds:
        if avg >= t:
            counts[t] += 1

for t, c in counts.items():
    print(f'Q>={t}: {c} reads')

Remove N Bases and Low Quality Together

python
def clean_read(record, min_qual=20):
    seq = str(record.seq)
    quals = record.letter_annotations['phred_quality']
    keep = [(s, q) for s, q in zip(seq, quals) if s != 'N' and q >= min_qual]
    if not keep:
        return None
    new_seq, new_quals = zip(*keep)
    new_record = record[:0]  # Empty copy with same metadata
    new_record.seq = Seq(''.join(new_seq))
    new_record.letter_annotations['phred_quality'] = list(new_quals)
    return new_record

FASTQ Format Variants

| Variant | Format String | Quality Encoding | ASCII Range | |---------|---------------|------------------|-------------| | Sanger/Illumina 1.8+ | 'fastq' | Phred+33 (standard) | 33-126 | | Solexa | 'fastq-solexa' | Solexa+64 | 59-126 | | Illumina 1.3-1.7 | 'fastq-illumina' | Phred+64 | 64-126 |

Most modern data uses standard 'fastq' (Sanger/Illumina 1.8+).

Quality Score Conversion

For legacy data using different quality encodings:

python
from Bio.SeqIO.QualityIO import phred_quality_from_solexa, solexa_quality_from_phred

Convert Solexa to Phred

python
from Bio.SeqIO.QualityIO import phred_quality_from_solexa

# Convert single score
solexa_score = 10
phred_score = phred_quality_from_solexa(solexa_score)

# Convert list of scores
solexa_scores = [10, 20, 30, 40]
phred_scores = [phred_quality_from_solexa(s) for s in solexa_scores]

Convert Phred to Solexa

python
from Bio.SeqIO.QualityIO import solexa_quality_from_phred

phred_score = 30
solexa_score = solexa_quality_from_phred(phred_score)

Convert Between FASTQ Variants

python
from Bio import SeqIO

# Read old Illumina format, write standard format
records = SeqIO.parse('old_reads.fastq', 'fastq-illumina')
SeqIO.write(records, 'standard_reads.fastq', 'fastq')

# Read Solexa format, write standard format
records = SeqIO.parse('solexa_reads.fastq', 'fastq-solexa')
SeqIO.write(records, 'standard_reads.fastq', 'fastq')

Auto-Detect Quality Encoding

Goal: Determine which FASTQ quality encoding (Sanger, Solexa, or Illumina 1.3+) a file uses.

Approach: Sample quality lines, find the minimum ASCII value, and compare against known offset ranges.

Reference (BioPython 1.83+):

python
def detect_quality_encoding(filepath, sample_size=1000):
    min_qual = 126
    max_qual = 0
    count = 0

    with open(filepath) as f:
        for i, line in enumerate(f):
            if i % 4 == 3:  # Quality line
                for char in line.strip():
                    min_qual = min(min_qual, ord(char))
                    max_qual = max(max_qual, ord(char))
                count += 1
                if count >= sample_size:
                    break

    if min_qual < 59:
        return 'fastq'  # Sanger/Illumina 1.8+ (Phred+33)
    elif min_qual < 64:
        return 'fastq-solexa'  # Solexa+64
    else:
        return 'fastq-illumina'  # Illumina 1.3+ (Phred+64)

Common Errors

| Error | Cause | Solution | |-------|-------|----------| | KeyError: 'phred_quality' | Not FASTQ or wrong variant | Check format, try 'fastq-illumina' | | Quality scores all 0 | Wrong encoding assumed | Try different FASTQ variant | | Trimmed reads empty | Too aggressive trimming | Lower quality threshold |

Related Skills

• read-sequences - Parse FASTQ files
• filter-sequences - Filter reads by other criteria (length, content)
• paired-end-fastq - Handle R1/R2 paired quality filtering
• sequence-statistics - Generate summary statistics including quality
• alignment-files - After filtering, align reads with bwa/bowtie2; quality scores in BAM