- Remote and local database searches are two important approaches used in bioinformatics to compare biological sequences against nucleotide, protein, and other biological databases. Both approaches help researchers identify sequence similarities, detect homologous genes and proteins, investigate evolutionary relationships, and predict biological functions. The main difference is where the database is stored and where the search computation takes place. In a remote search, the query is submitted to a server that accesses a database and performs the analysis. In a local search, the database is stored on a researcher’s own computer, workstation, or computing cluster, and the search is performed using locally installed software.
- Remote database searching is commonly used when researchers want to access large public biological databases without downloading and maintaining them. For example, the NCBI BLAST website allows users to submit DNA, RNA, or protein sequences and search against databases hosted by the National Center for Biotechnology Information. The query is transmitted to the remote service, the server performs the search, and the results are returned through a web interface or supported programmatic service. This approach is convenient for exploratory analyses, occasional sequence comparisons, and searches against databases that would require substantial storage and computing resources to maintain locally.
- Remote searches are available through several biological databases and online services. NCBI BLAST provides access to nucleotide and protein databases, including collections such as the nucleotide database and the non-redundant protein database. Ensembl provides genome and comparative-genomics resources, while UniProt offers protein sequence and functional information. Other services provide access to protein families, conserved domains, pathways, structures, and genetic variation. Researchers can use web interfaces or application programming interfaces (APIs), depending on the resource and the type of analysis.
- One major advantage of remote searching is that it reduces the need for local database installation and maintenance. Large sequence databases may require substantial disk space and frequent updates. By using a remote service, researchers can access databases maintained by specialist organizations without downloading all their contents. Remote searches are also useful when working on computers with limited memory, processing power, or storage capacity. The search server performs most of the computational work, allowing the user’s computer to focus on preparing queries and examining results.
- Remote searches can also provide access to specialized databases and filtering options that may be difficult to reproduce locally. For example, NCBI BLAST supports searches against available public databases and provides options for restricting searches using organism or taxonomic information. Researchers can use these features to investigate sequences from particular organisms or groups without constructing a separate local database. However, the exact search options depend on the remote service and database selected.
- Remote searching also has limitations. It requires a reliable internet connection, and performance depends on network speed, server availability, database size, and the number of other users submitting searches. Public services are shared resources, so large numbers of searches or high-volume automated requests may be restricted, delayed, or placed in lower-priority queues. Remote searching may therefore be unsuitable for projects involving millions of sequences or repeated large-scale comparisons. NCBI specifically recommends standalone BLAST against local databases or suitable cloud-based resources for large batch-search projects rather than relying on its public web service for extensive workloads. <Cite refs={[“turn264893search0″,”turn264893search10”]}/>
- Another consideration is data privacy. A remote search requires the query sequence to be transmitted to an external service. This may be acceptable for publicly available sequences, but it can be inappropriate for confidential, unpublished, proprietary, or otherwise sensitive biological data. Researchers working with human genomic data, clinical samples, commercial research, or unpublished discoveries should review the relevant data-sharing policies and institutional requirements before submitting sequences to external servers. Local searching can provide greater control over where the query data are processed and stored.
- Local database searching involves downloading or obtaining biological sequence data, preparing the database in a format supported by the selected software, and performing searches on a local computer or computing environment. The database may contain public sequences downloaded from resources such as NCBI or UniProt, or it may be a customized collection of sequences generated by a research project. Local searches are commonly performed using standalone software such as NCBI BLAST+, DIAMOND, MMseqs2, HMMER, and other sequence-analysis programs.
- A local search provides greater control over the database contents, search parameters, computational resources, and analysis workflow. Researchers can construct databases containing only sequences from selected organisms, specific gene families, particular experimental datasets, or a custom collection of proteins. This is especially useful when studying non-model organisms, newly sequenced genomes, microbial communities, or specialized biological datasets that are not adequately represented in general public databases.
- One important advantage of local searching is that it supports large-scale and repeated analyses. Once a database has been downloaded and prepared, researchers can perform many searches without repeatedly transmitting queries to an external server. Local searches can also be integrated into automated pipelines, shell scripts, Python programs, R workflows, and workflow-management systems. This makes them suitable for genome annotation, metagenomic classification, comparative genomics, transcriptome analysis, and large-scale protein function prediction.
- Local searches can be performed without a continuous internet connection once the software and required databases have been installed. This is useful in laboratories with limited network access or in environments where data must remain within a controlled computing infrastructure. Researchers can also retain a fixed database version, allowing analyses to be repeated using the same sequence collection rather than relying on a remote database that may have changed since an earlier search.
- The main limitations of local searching are the storage, installation, computational, and maintenance requirements. Large nucleotide and protein databases can occupy substantial disk space, and searching them may require significant memory, processing power, and storage performance. Researchers must install the appropriate software, download or construct the databases, manage updates, and ensure that the database files remain accessible and valid. Large-scale searches may require a workstation, high-performance computing cluster, or cloud infrastructure.
- The NCBI BLAST+ package provides standalone command-line programs for performing local searches and can also send searches to NCBI servers through its remote-search option. These are two different operating modes of the same software package. A local search uses a database installed on the user’s system, while a remote search sends the request to NCBI and uses its hosted databases and computing resources. <Cite refs={[“turn264893search5″,”turn264893search12”]}/>
- Before performing a local BLAST search, researchers must prepare the database in a format suitable for the selected BLAST program. The makeblastdb utility converts FASTA sequence files into BLAST database files. For a protein database, the -dbtype prot option is used:
makeblastdb \
-in protein_sequences.fasta \
-dbtype prot \
-parse_seqids \
-out local_protein_db
For a nucleotide database, the -dbtype nucl option is used:
makeblastdb \
-in nucleotide_sequences.fasta \
-dbtype nucl \
-parse_seqids \
-out local_nucleotide_db
- After database preparation, a protein query can be searched against the local protein database using BLASTP:
blastp \
-query query_protein.fasta \
-db local_protein_db \
-evalue 1e-5 \
-out protein_results.tsv \
-outfmt “6 qseqid sseqid pident length evalue bitscore qcovs”
Similarly, a nucleotide query can be searched against a local nucleotide database using BLASTN:
blastn \
-query query_nucleotide.fasta \
-db local_nucleotide_db \
-evalue 1e-10 \
-out nucleotide_results.tsv \
-outfmt “6 qseqid sseqid pident length evalue bitscore qcovs”
- These examples demonstrate local database searches using the BLAST+ command-line tools. The query and database must be compatible with the selected search program. For example, BLASTP requires a protein query and protein database, whereas BLASTN requires nucleotide sequences. Other programs, including BLASTX, TBLASTN, and TBLASTX, support searches involving translated nucleotide sequences.
- Remote searching can also be performed from the BLAST+ command line. In this approach, the local BLAST executable acts as a client and sends the query to the NCBI remote service. The -remote option instructs the program to use NCBI’s computing resources rather than a locally installed database. For example:
blastp \
-query query_protein.fasta \
-db nr \
-remote \
-evalue 1e-5 \
-out remote_results.tsv \
-outfmt “6 qseqid sseqid pident length evalue bitscore qcovs”
For a nucleotide search:
blastn \
-query query_nucleotide.fasta \
-db nt \
-remote \
-evalue 1e-10 \
-out remote_results.tsv \
-outfmt “6 qseqid sseqid pident length evalue bitscore qcovs”
- In these examples, nr and nt refer to NCBI-hosted protein and nucleotide database collections available for remote BLAST searches. The exact database names and supported options should be checked against the current NCBI service. The -remote option does not download the database to the local computer; it submits the search to the remote service. <Cite refs={[“turn264893search4″,”turn264893search5”]}/>
- Remote database searching can also be performed through web-based interfaces. A researcher can open the NCBI BLAST website, select the appropriate search program, paste a sequence or upload a FASTA file, choose a database, configure search parameters, and submit the query. The results are displayed through the browser and may include descriptions of matching sequences, alignment statistics, graphical representations, and detailed pairwise alignments. Web-based searches are convenient for occasional analyses and users who prefer graphical interfaces rather than command-line tools.
- Programmatic remote searches are possible through APIs and web services. For example, NCBI provides the BLAST URL API, which allows applications to submit searches and retrieve results programmatically. A typical workflow involves submitting a query, receiving a request identifier, checking whether the search has completed, and retrieving the results. Because the service is shared, applications must follow the provider’s usage guidelines, including appropriate request intervals and result polling. Large-scale automated searches should use suitable local or cloud-based infrastructure rather than repeatedly submitting individual requests to a public service. <Cite refs={[“turn264893search10”]}/>
- Local and remote searches can produce the same or very similar results when the same software version, database content, search parameters, and statistical settings are used. However, results may differ if the remote and local databases contain different sequences, if one database has been updated, or if the search programs use different versions or default parameters. Database size and composition can also influence E-values and the number of reported hits. Consequently, researchers should not assume that a search against a current remote database will reproduce an earlier search against an older local database.
- Database updates are especially important in local searching. Public biological databases are continually expanded and revised as new sequences are submitted, genome assemblies are improved, and annotations are corrected. A local database may become outdated if it is not refreshed. Researchers should establish an update schedule appropriate to the project and retain information about each database release or download date. For reproducible work, it is often preferable to preserve the exact database snapshot used in an analysis rather than replacing it without documentation.
- The choice between remote and local searching depends on the research objective, the number of queries, the size of the database, available computing resources, network access, privacy requirements, and the need for reproducibility. Remote searches are generally convenient for occasional queries against large public databases, exploratory analysis, and situations where local storage and computing resources are limited. Local searches are generally more appropriate for repeated analyses, large-scale projects, custom databases, automated pipelines, confidential data, and studies requiring a fixed database version.
- The two approaches can also be combined in a single workflow. Researchers may initially perform a remote search to identify candidate homologues or discover relevant database records. They can then download selected sequences and create a local database for more detailed comparisons. Alternatively, a local search can be used for routine analysis, followed by a remote search to investigate newly discovered sequences against a broader public collection. This combined strategy provides flexibility while allowing researchers to control the scale and reproducibility of their analyses.
- Local and remote database searches are widely used in genome annotation, protein identification, gene discovery, evolutionary analysis, metagenomics, pathogen identification, transcriptome studies, and comparative genomics. For example, a researcher may use remote BLAST to identify a protein’s likely family and then construct a local database containing related proteins from several organisms. An all-against-all local comparison can subsequently be used to identify gene families, examine conserved regions, and investigate orthologous and paralogous relationships.
- For reliable analysis, researchers should record the search program and version, database name and release or download date, query sequence, search parameters, filtering options, and output format. For remote searches, the service and relevant search identifiers should also be recorded where available. For local searches, the database source, construction commands, sequence files, and file checksums can be retained. These details help ensure that search results can be verified, reproduced, and interpreted correctly.
- Overall, remote and local database searches provide complementary approaches to biological sequence analysis. Remote searching offers convenient access to large, externally maintained databases without requiring local installation, while local searching provides greater control over data, computational workflows, custom databases, and reproducibility. Choosing the appropriate approach depends on the scale, purpose, and constraints of the research project. Understanding both methods allows bioinformatics researchers to perform efficient and reliable sequence searches across a wide range of biological applications.