Busco Genomics

BUSCO (Benchmarking Universal Single Copy Orthologs) is a bioinformatics tool that measures the completeness of genome and transcriptome assemblies by searching for a set of highly conserved single copy genes expected to be present in nearly all species of a given lineage. This guide provides a practical, source bounded framework for understanding BUSCO, deciding when to use it, implementing a robust workflow, and avoiding common pitfalls. It is intended for bioinformaticians, genomics researchers, and students who need to evaluate assembly quality or use ortholog sets for evolutionary analyses. The Galaxy Training Network offers hands on tutorials for BUSCO workflows Galaxy Training Network.

BUSCO works by comparing your assembly against a curated collection of lineage specific ortholog groups. If the majority of expected single copy orthologs are present and complete, the assembly is considered high quality. Several recent studies have adopted BUSCO as a standard metric, such as the de novo assembly of five Heliconius butterfly genomes which reported high BUSCO scores to validate contiguity de novo assembly of 5 highly contiguous Heliconius butterfly genomes. Similarly, the chromosome level genome of a snail Sinotaia quadrata used BUSCO to confirm assembly completeness Chromosome level genome assembly of a snail Sinotaia quadrata.

What is BUSCO Genomics

BUSCO stands for Benchmarking Universal Single Copy Orthologs. It was developed to assess assembly completeness in a way that is independent of assembly size or N50. The method relies on lineage specific ortholog databases, such as those for eukaryotes, arthropods, vertebrates, or plants. These databases contain sets of genes that are present as single copies in nearly all species of the lineage. If a genome assembly contains most of these genes in a complete state, it suggests the assembly captured the majority of the gene space. The EMBL EBI training resources describe how BUSCO uses these lineage datasets to produce a standardized quality score EMBL EBI Training. The tool outputs four categories: complete single copy, complete duplicated, fragmented, and missing. For a typical high quality genome, you expect more than 90 percent complete BUSCOs.

At a Glance

When to Use BUSCO

Deciding when to apply BUSCO depends on your goal and data type. Use BUSCO when you need to:

• Assess the completeness of a new genome assembly before publication. Many journals now require BUSCO scores, as seen in the chromosome level genome of Yarra pygmy perch Chromosome level genome assemblies of Southern pygmy perch and Yarra pygmy perch.
• Compare the quality of different assemblies produced by different assemblers or sequencing technologies.
• Validate transcriptome assemblies from RNA seq data, especially when gene space coverage is important.
• Extract conserved orthologs for phylogenomic analyses. The fish parasite Trypanosoma larimichthysi genome used BUSCO orthologs to place the species taxonomically A chromosome level gap free genome assembly of the fish parasite Trypanosoma larimichthysi.
• Guide annotation efforts by identifying regions that may be missing or misassembled.

Do not use BUSCO if you need to assess noncoding regions, structural rearrangements, or repetitive elements. BUSCO only targets conserved protein coding genes.

Practical Workflow

A typical BUSCO workflow involves these steps:

1. Install BUSCO. Download the latest version from the GitHub repository or use a package manager like conda. The Galaxy platform also provides a point and click interface Galaxy Training Network.

2. Choose the correct lineage dataset. BUSCO provides multiple lineages (e.g., eukaryota_odb10, arthropoda_odb10, vertebrata_odb10). Select the most specific lineage that matches your organism. Using a lineage that is too broad may reduce sensitivity, using one too narrow may miss orthologs.

3. Run BUSCO with appropriate flags.

   • -i input assembly (FASTA)
   • -l lineage dataset
   • -o output directory
   • -m mode: genome, transcriptome, or proteins For a genome assembly use -m genome. For a transcriptome set -m transcriptome. The transcriptome dataset of seven nerites relied on transcriptome mode to evaluate their de novo assembly Transcriptome dataset of seven nerites.

4. Review the output. BUSCO generates a short summary text file and a full table with ortholog status. The summary shows the number and percentage of complete, fragmented, and missing BUSCOs. A run that yields over 90 percent complete single copy BUSCOs is generally considered good for a genome assembly.

5. Interpret the results. Compare your scores to published benchmarks for your taxon. For example, the Pangasius nasutus genome reported a BUSCO completeness of 94.5 percent using the actinopterygii lineage Genomic insights from the Malaysian Pangasiid. Scores below 80 percent may indicate a fragmented assembly, poor sequencing depth, or contamination.

Quality Checks and Interpretation

Always check that your BUSCO run completed without errors. Look at the log file for warnings. Verify that the lineage dataset matches your organism. For genomes with high heterozygosity or polyploidy, the proportion of duplicated BUSCOs may be elevated. A high duplicate count can indicate a collapsed assembly of alleles or a true duplication event. In such cases, compare with a closely related reference. For transcriptomes, a higher missing rate is acceptable because not all genes are expressed in all tissues. The Galaxy Training Network provides tutorials for interpreting these scores Galaxy Training Network. Cross check BUSCO results with other completeness metrics like N50 or L50, but remember that BUSCO evaluates gene content, not contiguity.

Common Mistakes

Using the wrong lineage dataset is the most frequent error. If you run a vertebrate genome with the eukaryota lineage, you may get inflated complete scores because the eukaryota set contains fewer and more conserved genes, masking fragmentation. Another mistake is running transcriptome mode on a genome assembly, which expects protein coding sequences and may miss many orthologs. Conversely, running genome mode on a transcriptome will treat each contig as a separate chromosome, leading to many duplicated hits. Some users also assume that a high BUSCO score alone guarantees a good assembly. But BUSCO does not detect chimeric contigs, misjoins, or missing noncoding regions. Additionally, poorly polished assemblies with frameshifts in conserved genes may be scored as fragmented. Always inspect a few representative loci.

Limits and Uncertainty

BUSCO has several inherent limitations. It only surveys single copy orthologs, which represent a small fraction of the genome (typically a few hundred to a few thousand genes). It provides no information about structural accuracy, repetitive sequences, or regulatory elements. Species that have undergone whole genome duplication or recent gene family expansions may show inflated duplicate counts that are not biological. The datasets are also biased toward well studied model organisms. For highly divergent lineages, the ortholog detection may fail because the genes have diverged beyond recognition. The NCBI Sequence Read Archive hosts many assemblies that rely on BUSCO for validation, but users should treat BUSCO as one of several quality metrics, not the sole arbiter NCBI Sequence Read Archive. In the Yarra pygmy perch study, BUSCO scores were complemented by synteny analysis to confirm chromosome level assembly Chromosome level genome assemblies of Southern pygmy perch and Yarra pygmy perch. Always consider your biological question. A BUSCO score of 95 percent may be excellent for draft assembly, but insufficient for fine scale evolutionary studies.

Frequently Asked Questions

1. Can I run BUSCO on a transcriptome assembly without a reference genome? Yes, use transcriptome mode (-m transcriptome). The tool will search the assembled transcripts for orthologs. However, completeness depends on the tissue sampling depth. Only expressed genes are detectable. For best results, include RNA from multiple tissues and conditions.

2. What does a high duplications percentage mean? A high duplication rate often indicates that the assembly contains multiple copies of the same ortholog, which can arise from allelic variation in a diploid genome, recent gene duplications, or collapsed repeats. Check if the percentage is consistent with the known biology of your organism. If it exceeds 5 to 10 percent, consider whether the assembly has not been haploidized properly.

3. Why did BUSCO return a very low completeness score for my genome? Low scores can be due to several reasons: poor sequencing coverage, a fragmented assembly, contamination from other species, or using an inappropriate lineage dataset. Verify your input assembly, rerun with a more general lineage, and examine the missing orthologs to see if they are clustered or scattered.

4. Is there a BUSCO alternative? Several tools exist, such as CEGMA (a predecessor), but BUSCO is the most widely adopted. For specific purposes, you might combine BUSCO with tools like QUAST for assembly metrics or with BlobTools for contamination analysis. The Bioconductor project offers packages for downstream analysis of genomic data, but BUSCO itself is a standalone tool Bioconductor.

References and Further Reading

• Galaxy Training Network. Practical BUSCO workflow tutorials. Galaxy Training Network
• EMBL EBI Training. Introduction to genome assembly evaluation. EMBL EBI Training
• NCBI Bookshelf. Genome assembly quality assessment. NCBI Bookshelf
• NCBI Sequence Read Archive. Public repository of BUSCO validated assemblies. NCBI Sequence Read Archive
• Bioconductor. Tools for genomic analysis. Bioconductor
• Transcriptome dataset of seven nerites (Clithon, Neripteron, and Nerita, Neritidae). Sci Data. Link
• de novo assembly of 5 highly contiguous Heliconius butterfly genomes with long read sequencing alone. Genome Biol Evol. Link
• Genomic insights from the Malaysian Pangasiid, Pangasius nasutus, reveal evolutionary divergence and functional adaptations in Pangasiid catfishes. Sci Rep. Link
• Chromosome level genome assembly of a snail Sinotaia quadrata (Gastropoda: Viviparidae). Sci Data. Link
• Chromosome level genome assemblies of Southern pygmy perch (Nannoperca australis) and Yarra pygmy perch (N. obscura) reveal several miniaturization genes. J Hered. Link
• A chromosome level, gap free genome assembly of the fish parasite Trypanosoma larimichthysi from large yellow croaker. Sci Data. Link

Related Articles

• Cell Cycle
• Dna Replication
• Protein Synthesis
• Incomplete Dominance Gene
• Cell Membrane Function Biology