PCRpanel: automated design of ultra-multiplex PCR panels for targeted amplicon sequencing

Le résumé fourni par la source

Ultra-multiplex PCR has become essential for targeted amplicon sequencing in molecular biology and clinical genetics. We implemented PCRpanel, a command-line Java application with a companion web interface that jointly optimises primer thermodynamics, linguistic sequence complexity, primer-dimer interactions, and multiplex-aware, reference-guided specificity and repeat masking for automated ultra-multiplex panel design. We benchmarked PCRpanel in silico against NGS-PrimerPlex and Olivar and validated it experimentally by designing tiled primer pools targeting all exonic regions of COL4A3, COL4A4, COL4A5 and COL4A6 and amplifying DNA from four clinical Alport syndrome samples. As this was a pilot run aimed at initial validation of the design, sequencing is reported for one representative patient sample, in which each primer pool was amplified separately and sequenced at two template dilutions, giving 14 libraries on the Illumina MiSeq platform; pool 1.1 was run as a 50-primer sub-pool during protocol optimisation. Reads were aligned to GRCh38 with the DRAGEN Bio-IT Platform v4.4.4 and coverage was quantified over an exon-restricted target interval of 45,372 bp derived from GENCODE v49. Duplicate marking was disabled, because amplicon molecules generated by a common primer pair are positionally indistinguishable from PCR duplicates. Here we present PCRpanel, a command-line Java application for the automated design of thermodynamically optimised primer panels for short- and long-read amplicon sequencing. PCRpanel supports workflows from simple two-primer assays to ultra-high-plex designs containing hundreds of amplicons and accommodates targets ranging from complete viral genomes and eukaryotic gene families to structural-variant breakpoints and environmental metagenomes. The software jointly optimises sequence complexity, thermodynamic stability, primer-dimer interactions, and off-target amplification, enabling robust primer selection even in repetitive or homologous regions. Efficient algorithms generate ultra-high-plex panels within seconds to minutes on standard hardware, up to 15 min when the full human reference genome is screened, and both gene-specific and universal designs are supported across homologous gene families. We validated the approach by designing 237 primer pairs (474 primers) targeting all exonic regions of the three established Alport syndrome genes COL4A3, COL4A4 and COL4A5, together with the adjacent COL4A6. Experimental validation on four clinical Alport syndrome DNA samples confirmed successful amplification of all 237 primer pairs. As this was a pilot run, sequencing of one representative patient sample, in which each primer pool was amplified separately and sequenced at two template dilutions as 14 libraries (pool 1.1 as a 50-primer sub-pool during protocol optimisation), was used to evaluate analytical performance: read alignment rates were 94.2–99.3%, 77.8–96.1% of aligned reads fell within the exonic target and mean target depth was 149.1-273.7x per library, with target bases that received amplicon coverage sequenced far above the thresholds required for genotype determination (97.9–100% of bases at ≥ 20x also reaching ≥ 50x). Because each library carried a single primer pool, coverage breadth normalised to the complete four-gene target reflects that pooling design rather than amplification performance; measured against the footprint each pool was designed to produce, the covered ≥ 1× footprint of each library corresponded to approximately 60–130%. In silico exonic coverage was 65.6% for COL4A3, 41.4% for COL4A4 and 64.4% for COL4A6. For COL4A5 the same settings covered 30.8% of exonic bases, under the amplicon-size window and genome-wide specificity screen used for the deposited panel; a systematic sweep of 1,688 design runs raised coverage to 92.8% of the exonic target in a single configuration with repeat masking enforced in full and the genome-wide specificity screen omitted, and placed an amplicon over all 53 exons across the sweep (Supplementary Table S6 b), so the shortfall reflects the parameters used for the deposited panel together with that screen rather than a limit of the method. A step-by-step gap-filling procedure, which restricts redesign to the uncovered intervals and merges rescue primers into the panel under dimer control, is provided for recovering such regions at a stated off-target risk. PCRpanel therefore provides a scalable, platform-agnostic solution for ultra-multiplex PCR panel construction, supporting accurate and reproducible targeted sequencing in both research and diagnostic settings.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.