--- name: bio-small-rna-seq-target-prediction description: Predict miRNA target genes using sequence-based algorithms and database lookups. Use when identifying potential mRNA targets of differentially expressed or functionally important miRNAs. tool_type: mixed primary_tool: miRanda measurable_outcome: Execute skill workflow successfully with valid output within 15 minutes. allowed-tools: - read_file - run_shell_command --- # miRNA Target Prediction ## miRanda Algorithm ```bash # Run miRanda for target prediction miranda miRNA.fa UTRs.fa \ -sc 140 \ -en -20 \ -out predictions.txt # Options: # -sc 140: Minimum alignment score (default 140) # -en -20: Maximum free energy threshold (kcal/mol) # Higher score and lower energy = stronger prediction ``` ## Parse miRanda Output ```python import pandas as pd def parse_miranda(output_file): '''Parse miRanda output file''' results = [] with open(output_file) as f: for line in f: if line.startswith('>'): parts = line.strip().split('\t') if len(parts) >= 5: results.append({ 'mirna': parts[0].lstrip('>'), 'target': parts[1], 'score': float(parts[2]), 'energy': float(parts[3]), 'position': parts[4] }) return pd.DataFrame(results) ``` ## TargetScan Database Lookup ```python import requests import pandas as pd def query_targetscan(mirna_family): '''Query TargetScan for predicted targets Note: TargetScan uses miRNA family names (e.g., miR-21-5p) ''' # TargetScan provides downloadable files # For human: https://www.targetscan.org/vert_80/vert_80_data_download/ targetscan_file = 'Predicted_Targets_Context_Scores.txt' df = pd.read_csv(targetscan_file, sep='\t') targets = df[df['miRNA family'] == mirna_family] return targets.sort_values('context++ score') ``` ## miRDB Database Lookup ```python def query_mirdb(mirna_id): '''Query miRDB for target predictions miRDB uses machine learning for target prediction Score > 80 indicates high confidence ''' # Download from http://mirdb.org/download.html mirdb_file = 'miRDB_v6.0_prediction_result.txt' df = pd.read_csv(mirdb_file, sep='\t', header=None, names=['mirna', 'target', 'score']) targets = df[df['mirna'] == mirna_id] return targets[targets['score'] >= 80].sort_values('score', ascending=False) ``` ## Combine Multiple Databases ```python def consensus_targets(mirna, min_databases=2): '''Find targets predicted by multiple databases More reliable targets are predicted by multiple algorithms ''' miranda_targets = set(query_miranda_targets(mirna)) targetscan_targets = set(query_targetscan_targets(mirna)) mirdb_targets = set(query_mirdb_targets(mirna)) # Count predictions per target all_targets = miranda_targets | targetscan_targets | mirdb_targets consensus = [] for target in all_targets: count = sum([ target in miranda_targets, target in targetscan_targets, target in mirdb_targets ]) if count >= min_databases: consensus.append({ 'target': target, 'n_databases': count, 'miranda': target in miranda_targets, 'targetscan': target in targetscan_targets, 'mirdb': target in mirdb_targets }) return pd.DataFrame(consensus).sort_values('n_databases', ascending=False) ``` ## Python miRNA Target Prediction ```python # Using mirtarbase package for validated targets def get_validated_targets(mirna): '''Get experimentally validated targets from miRTarBase''' # Download from https://mirtarbase.cuhk.edu.cn/ mirtarbase_file = 'miRTarBase_MTI.xlsx' df = pd.read_excel(mirtarbase_file) validated = df[df['miRNA'] == mirna] return validated[['Target Gene', 'Experiments', 'Support Type']] ``` ## Seed Match Analysis ```python from Bio.Seq import Seq def find_seed_matches(mirna_seq, utr_seq): '''Find seed matches in UTR sequence Seed region: positions 2-8 of miRNA (7-mer) ''' mirna = Seq(mirna_seq) utr = Seq(utr_seq) # Get seed (positions 2-8, 0-indexed: 1-7) seed = str(mirna[1:8]) seed_rc = str(Seq(seed).reverse_complement()) matches = [] start = 0 while True: pos = str(utr).find(seed_rc, start) if pos == -1: break matches.append(pos) start = pos + 1 return matches ``` ## Functional Enrichment of Targets ```python def enrich_target_genes(targets, background=None): '''Run GO enrichment on predicted target genes''' import gseapy as gp enr = gp.enrichr( gene_list=targets, gene_sets=['GO_Biological_Process_2021', 'KEGG_2021_Human'], organism='Human' ) return enr.results ``` ## Related Skills - differential-mirna - Get DE miRNAs for target prediction - pathway-analysis/go-enrichment - Enrich target gene functions - database-access/entrez-fetch - Query biological databases --- Source attribution: GPTomics Original project: https://github.com/GPTomics/bioSkills License: MIT License information: https://opensource.org/license/mit Original upstream instructions; downloaded from Mega AI Hub. ## Upstream license notice MIT License Copyright (c) 2026 FridrichMethod Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions: The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software. THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. --- NOTE ON BUNDLED CONTENT This repository aggregates skills from multiple upstream projects (see the README for the source list). Each `skills//` directory may carry its own license inherited from its upstream source. The MIT license above applies to this repository's curation and the scripts/workflows in `/`, `/scripts/`, and `/.github/` — not to the upstream skill content itself. Consult the upstream repository linked in the README and any LICENSE file inside a skill directory before reusing skill content commercially. # Mega AI Hub resource metadata Creator / publisher: GPTomics Original project: https://github.com/GPTomics/bioSkills Attribution basis: Exact package in biomedical source index and publisher source table Fields of work: Data & Analytics, Education & Research Roles: Business Intelligence Analyst, Data Scientist, Researcher Specialties: Backend, Database, Excel, Frontend Tools: Targetscan, Mirdb, Miranda Country applicability: Country not specified Classification basis: Source catalog labels Instruction SHA-256: c001cebc4899e760a2e7c8d9229ec6f6b484f5bbdd1a27699b263e81a9c6dd2b Metadata reviewed: 2026-10-27 This metadata supersedes earlier Mega AI Hub catalog attribution. Original instruction and supporting files are preserved.