# Bioinformatics > Computational biology tools including sequence alignment, phylogenetic analysis, genome assembly, variant calling, and biological database queries for computational research. - Skill: `neuralblitz/bioinformatics-3` (Agent Skill) - Install (CLI): `npx skillmds@latest add neuralblitz/bioinformatics-3` - Raw SKILL.md: https://api.skillmd.com/api/skills/neuralblitz/bioinformatics-3/raw - Safety review: pending (external: skill-scanner PASS, skillspector PASS) - Works with: Claude Code, Claude.ai, OpenAI Codex - Category: AI & ML - License: MIT - Author: NeuralBlitz (https://skillmd.com/u/neuralblitz) - Updated: 2026-09-22 - Page: https://skillmd.com/skills/neuralblitz/bioinformatics-3 --- # Bioinformatics ## What I Do I provide comprehensive bioinformatics tools including sequence alignment algorithms, phylogenetic analysis, genome assembly, variant calling, biological database queries, and structural bioinformatics for computational biology applications. ## When to Use Me - DNA/Protein sequence alignment - Phylogenetic tree construction - Genome assembly and annotation - Variant calling and annotation - Biological database mining - Protein structure analysis ## Core Concepts - **Sequence Alignment**: Global, local, pairwise, multiple - **Substitution Matrices**: BLOSUM, PAM - **Phylogenetics**: Distance, character, tree building - **Genome Assembly**: Overlap-layout-consensus, de Bruijn - **Variant Calling**: SNPs, indels, structural variants - **Machine Learning**: Protein function prediction - **Structural Bioinformatics**: Homology modeling, docking - **Databases**: NCBI, UniProt, PDB, Ensembl ## Code Examples ### Sequence Alignment ```python from collections import defaultdict BLOSUM62 = { ('A', 'A'): 4, ('A', 'C'): 0, ('A', 'D'): -2, ('A', 'E'): -1, ('A', 'R'): -1, ('A', 'N'): -2, ('A', 'K'): -1, ('A', 'M'): -1, ('A', 'F'): -2, ('A', 'S'): 1, ('A', 'T'): 0, ('A', 'W'): -3, ('A', 'Y'): -2, ('A', 'V'): 0 } def needleman_wunsch(seq1, seq2, gap_penalty=-2): m, n = len(seq1), len(seq2) dp = [[0] * (n + 1) for _ in range(m + 1)] for i in range(m + 1): dp[i][0] = i * gap_penalty for j in range(n + 1): dp[0][j] = j * gap_penalty for i in range(1, m + 1): for j in range(1, n + 1): match = dp[i-1][j-1] + BLOSUM62.get((seq1[i-1], seq2[j-1]), 0) delete = dp[i-1][j] + gap_penalty insert = dp[i][j-1] + gap_penalty dp[i][j] = max(match, delete, insert) return dp[m][n] def smith_waterman(seq1, seq2, gap_penalty=-1): m, n = len(seq1), len(seq2) dp = [[0] * (n + 1) for _ in range(m + 1)] for i in range(1, m + 1): for j in range(1, n + 1): match = dp[i-1][j-1] + BLOSUM62.get((seq1[i-1], seq2[j-1]), 0) delete = dp[i-1][j] + gap_penalty insert = dp[i][j-1] + gap_penalty dp[i][j] = max(0, match, delete, insert) return max(max(row) for row in dp) score = needleman_wunsch("MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFLSFPTTKTYFPHFDLSH", "MVLSAADKTNVKAAWSKVGGHAGEYGAEALERMFLGFPTTKTYFPHFDLSH") print(f"Global alignment score: {score}") ``` ### Phylogenetic Analysis ```python from scipy.cluster.hierarchy import linkage, dendrogram from scipy.spatial.distance import pdist def calculate_distance_matrix(sequences): n = len(sequences) dist_matrix = np.zeros((n, n)) for i in range(n): for j in range(i + 1, n): dist = 1 - needleman_wunsch(sequences[i], sequences[j]) / max(len(sequences[i]), len(sequences[j])) dist_matrix[i, j] = dist_matrix[j, i] = dist return dist_matrix def upgma(distance_matrix): n = len(distance_matrix) clusters = {i: [i] for i in range(n)} heights = {i: 0 for i in range(n)} for step in range(n - 1): min_dist = float('inf') merge = (0, 0) for i in range(n): for j in range(i + 1, n): if distance_matrix[i, j] < min_dist and i in clusters and j in clusters: min_dist = distance_matrix[i, j] merge = (i, j) new_height = max(heights[merge[0]], heights[merge[1]]) + min_dist / 2 clusters[merge[0]].extend(clusters[merge[1]]) heights[len(clusters)] = new_height del clusters[merge[1]] return clusters seqs = ["SEQ1", "SEQ2", "SEQ3", "SEQ4"] dists = calculate_distance_matrix(seqs) tree = upgma(dists) print("UPGMA tree constructed") ``` ### Variant Calling Basics ```python def calculate_phred_quality(accuracy): return -10 * np.log10(1 - accuracy) def fastq_to_fasta(input_file, output_file): with open(input_file, 'r') as f_in, open(output_file, 'w') as f_out: while True: header = f_in.readline().strip() if not header: break seq = f_in.readline().strip() plus = f_in.readline() qual = f_in.readline().strip() f_out.write(f">{header[1:]}\n{seq}\n") def vcf_header(sample_name): return f"""##fileformat=VCFv4.2 ##source=MyVariantCaller ##sample={sample_name} ##INFO= ##FORMAT= #CHROM POS ID REF ALT QUAL FILTER INFO FORMAT SAMPLE """ def parse_vcf_line(line): fields = line.strip().split('\t') return { 'chrom': fields[0], 'pos': int(fields[1]), 'ref': fields[3], 'alt': fields[4], 'qual': float(fields[5]) if fields[5] != '.' else None, 'genotype': fields[9].split(':')[0] } ``` ### Protein Analysis ```python from collections import Counter AMINO_ACID_PROPERTIES = { 'hydrophobic': set('AILMFVWP'), 'hydrophilic': set('RNDCEQGHKSTY'), 'positively_charged': set('RKH'), 'negatively_charged': set('DE'), 'aromatic': set('FWY') } def calculate_protein_properties(sequence): aa_counts = Counter(sequence) length = len(sequence) properties = { 'length': length, 'molecular_weight': calculate_mw(sequence), 'isoelectric_point': predict_pi(sequence), 'hydrophobic_ratio': sum(aa_counts[aa] for aa in 'AILMFVWP') / length, 'charge_at_ph7': net_charge(sequence, 7) } return properties def calculate_mw(sequence): mw = 0 for aa in sequence: mw += AA_WEIGHTS.get(aa, 110) return mw - 18 * (len(sequence) - 1) def predict_pi(sequence): pKa = {'D': 3.9, 'E': 4.3, 'C': 8.3, 'Y': 10.1, 'H': 6.0, 'K': 10.5, 'R': 12.5} charges = {ph: 0 for ph in [5, 6, 7, 8, 9, 10]} for aa in sequence: for ph in charges: if aa in pKa: if aa in ['D', 'E', 'C', 'Y']: charges[ph] -= 1 / (1 + 10**(ph - pKa[aa])) else: charges[ph] += 1 / (1 + 10**(pKa[aa] - ph)) return min(range(5, 11), key=lambda x: abs(charges[x])) def find_membrane_regions(sequence, window=20): hydrophobicity = {'A': 1.8, 'C': 2.5, 'D': -3.5, 'E': -3.5, 'F': 2.8, 'G': -0.4, 'H': -3.2, 'I': 4.5, 'K': -3.9, 'L': 3.8, 'M': 1.9, 'N': -3.5, 'P': -1.6, 'Q': -3.5, 'R': -4.5, 'S': -0.8, 'T': -0.7, 'V': 4.2, 'W': -0.9, 'Y': -1.3} regions = [] for i in range(len(sequence) - window): avg = np.mean([hydrophobicity.get(aa, 0) for aa in sequence[i:i+window]]) if avg > 1.0: regions.append((i, i+window, avg)) return regions ``` ### Database Queries ```python import urllib.request import json def query_uniprot(accession): url = f"https://rest.uniprot.org/uniprotkb/{accession}.json" with urllib.request.urlopen(url) as response: return json.loads(response.read()) def query_ncbi_blast(query_sequence, database='nr', max_results=10): from Bio.Blast import NCBIWWW result_handle = NCBIWWW.qblast('blastp', database, query_sequence) return result_handle.read() def query_ensembl(gene_id): url = f"https://rest.ensembl.org/lookup/id/{gene_id}?expand=1" with urllib.request.urlopen(url) as response: return json.loads(response.read()) def calculate_sequence_identity(seq1, seq2): alignments = needleman_wunsch(seq1, seq2) return alignments / max(len(seq1), len(seq2)) ``` ## Best Practices 1. **Quality Control**: Filter low-quality sequences 2. **Reference Bias**: Use appropriate reference genomes 3. **Multiple Testing**: Correct for statistical testing 4. **Data Formats**: Use standard formats (FASTA, VCF, BAM) 5. **Reproducibility**: Document all parameters ## Common Patterns ```python # K-mer counting def count_kmers(sequence, k): kmers = {} for i in range(len(sequence) - k + 1): kmer = sequence[i:i+k] kmers[kmer] = kmers.get(kmer, 0) + 1 return kmers # Sliding window analysis def sliding_window_analysis(sequence, window_size, analysis_func): results = [] for i in range(len(sequence) - window_size + 1): window = sequence[i:i+window_size] results.append(analysis_func(window)) return results ``` ## Core Competencies 1. Sequence alignment algorithms 2. Phylogenetic tree construction 3. Variant calling and annotation 4. Protein sequence analysis 5. Biological database integration