Search results from local alignment search tools use statistical scores that are sensitive to the size of the database to report the quality of the result. For example, NCBI BLAST reports the best matches using similarity scores and expect values (i.e., e-values) calculated against the database size. Given the astronomical growth in genomics data throughout a genomic research investigation, sequence databases grow as new sequences are continuously being added to these databases. As a consequence, the results (e.g., best hits) and associated statistics (e.g., e-values) for a specific set of queries may change over the course of a genomic investigation. Thus, to update the results of a previously conducted BLAST search to find the best matches on an updated database, scientists must currently rerun the BLAST search against the entire updated database, which translates into irrecoverable and, in turn, wasted execution time, money, and computational resources. To address this issue, we devise a novel and efficient method to redeem past BLAST searches by introducing iBLAST. iBLAST leverages previous BLAST search results to conduct the same query search but only on the incremental (i.e., newly added) part of the database, recomputes the associated critical statistics such as e-values, and combines these results to produce updated search results. Our experimental results and fidelity analyses show that iBLAST delivers search results that are identical to NCBI BLAST at a substantially reduced computational cost, i.e., iBLAST performs (1 + δ)/δ times faster than NCBI BLAST, where δ represents the fraction of database growth. We then present three different use cases to demonstrate that iBLAST can enable efficient biological discovery at a much faster speed with a substantially reduced computational cost. For example, NCBI BLAST reports the best matches using similarity scores and expect values (i.e., e-values) calculated against the database size. Given the astronomical growth in genomics data throughout a genomic research investigation, sequence databases grow as new sequences are continuously being added to these databases. As a consequence, the results (e.g., best hits) and associated statistics (e.g., e-values) for a specific set of queries may change over the course of a genomic investigation. In reality, genome sequencing and annotation projects can be fairly long term, and thus, can require multiple sequence updates, e.g., regular annotation updates [ 12 , 13 ]. However, such updates require executing sequence similarity search from scratch as BLAST uses similarity scores and e-values that depend on the ever-increasing size of the database. For this reason, it is currently required to discard the results of prior search efforts and BLAST concepts and statistics Core concepts of a BLAST result: Hit, HSP, score, and e-value When we perform a BLAST search against a sequence database with a query sequence, the BLAST program returns the sequences producing significant alignment from the target database, which we refer to as hits . Between the query and a hit sequence, there exist many pairwise locally-optimal gapped local alignments, which we refer to as high scoring pairs or HSPs . The definitions of hits and hsps are slightly different from those used by Althshul and colleagues [ 14 ], but follows the structural definition from the XML output format produced by NCBI BLAST. First, the search is only run on the databases constructed from new sequences that have been added to the database. Second, the results generated from searching the new sequences in the database are then merged with the saved results from the previous BLAST search. e-value correction for Karlin-Altschul statistics Let n c represent the current database length and n d represent the length of the newly arrived sequences for the database. Also, let N c be the number of sequences in the current database and N d be the number of sequences in the newly arrived part of the database. Then, we have Actual length of the updated database : n t = n c + n d . S3 Algorithm (in S1 File ) documents the procedure to merge the hits from two results for the same query. All statistical parameters dependent on total database size are re-calibrated to recompute or re-scale the e-values. The hits are selected in the ascending order of their e-values (descending order of their scores). Additional details on recomputing and re-scaling e-values is provided in the S1 File (Section “e-value correction.”). iBLAST implementation We develop iBLAST for performing BLAST search as an extension to the NCBI BLAST code. It consists of Python wrapper scripts around the extended BLAST code and uses NCBI BLAST programs as black-box routines. Although this number is low, the identity of high-scoring hits is likely to be similar even if the best taxonomic hit to a query sequence was not retrieved, as gall wasps do not have any close relatives currently hosted in NCBI databases but rather many equidistant relatives. Given this, we then added in sequences of the rest of Hymenopteran species to see if this improves the number of shared top hits. With this analysis, we conducted BLAST search on only 1.17% of the total nr yet obtained 87.75% similarity in top hits to a full nr BLAST. This result demonstrates the potential of performing more taxon-guided approaches to save on the costs of large-scale BLAST searching jobs. Performing the analysis in this way has also enabled improved curation of hits by taxon, which facilitates better biological interpretation of these results. 10.1371/journal.pone.0249410.t003 Table 3 Potential for taxon-guided searches enabled by iBLAST. Comparison of merged BLAST results from multiple individual BLAST searches with a separate BLAST search conducted against a completed nr database shows that biologically relevant taxa can be added incrementally to obtain similar results to nr by searching against a much smaller database size.