Abstract Accurate methylation sequencing from clinical samples is challenging due to DNA-damaging chemistries that reduce sequence diversity, limiting data quality and reproducibility—particularly for fragmented or low-input materials such as cfDNA. The objective of this study was to evaluate a new positive-readout methylation method (TAPS+) in a clinically relevant cohort and compare its performance to an enzymatic methylation conversion method employing a negative-readout approach. TAPS+ is a non-damaging chemistry that converts methylated cytosines to thymines while preserving unmethylated cytosines, maintaining full sequence complexity and enabling direct methylation detection. This method was applied to a 24-sample cohort comprising invasive breast carcinoma patients, healthy donors, and control samples spanning a range of methylation states. Libraries were sequenced in parallel with those from an enzymatic negative-readout method to assess conversion efficiency, background, and biological concordance of differentially methylated regions (DMRs). The R package methylKit was used for DMR analysis, and biomarker candidates were identified based on statistically significant methylation differences within annotated genomic regions. Compared to the negative-readout method, TAPS+ achieved a shorter turnaround time and higher library yields across all sample types. In cfDNA samples, TAPS+ produced significantly higher methylation ratios than the negative-readout method (p = 8.6e-10). A methylated control showed an 88.8% 5mC rate with TAPS+, versus 76.8% with the negative-readout method, suggesting greater accuracy. CpG island methylation was strongly correlated across methods, though the negative-readout method had greater noise at low coverage sites. Both breast cancer patients and healthy individuals exhibited a characteristic dip in average DNA methylation levels around transcription start sites, consistent with promoter hypomethylation, with no significant differences observed between cohorts. DMR analysis with TAPS+ identified significant biomarker candidates, including SHH hypermethylation (q = 3.15e-5), associated with cancer development and progression. Additional biomarkers were associated with developmental pathways, cancer signaling, transcriptional regulation, and immune response. The negative-readout dataset did not demonstrate strong biomarker candidates indicative of disease status. TAPS+ enables high-fidelity methylation sequencing from diverse and challenging clinical materials, supporting robust epigenetic analysis within a streamlined workflow. The positive-readout approach demonstrated superior data quality and biological relevance compared to an enzymatic negative-readout method, highlighting its potential utility for translational research, early detection, and clinical assay development. Citation Format: Kimberly A. Holden, Kerry D. Fitzgerald, Adib Shafi, Ashraf Shabaneh, Dennis D. Krutkin, Tong Liu, Eyad Almasri, Graham McLennan, Nathan Faulkner, Craig Marshall, Travis Sanders, Thomas Harrison, Eduard Casas, Kristina Giorda, Doug Wendel, Brian Kudlow, Shakti Ramkissoon, Marcia Eisenberg, Brian Caveney, Eric Severson, Taylor J. Jensen, Jonathan Williams. Utility of TAPS+: a positive-readout methylation sequencing approach for high-fidelity epigenetic profiling [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 3213.

Abstract Comprehensive tumor characterization increasingly requires both genomic and epigenomic information, yet current methylation profiling methods, such as bisulfite-based workflows, limit sensitivity and accuracy due to DNA degradation and reduced sequence complexity. To address these constraints, we developed an improved positive-conversion chemistry that directly converts methylated cytosines while preserving unmethylated ones to enable simultaneous detection of cytosine modifications and genetic variants from a single NGS library. This approach is designed to support oncology research applications where high accuracy from limited or damaged material, such as FFPE tissue and cell-free circulating tumor DNA (ctDNA), is essential. To assess performance, FFPE, fresh-frozen, and cfDNA samples were processed using an optimized oxidation–reduction workflow. Libraries were sequenced on standard short-read platforms, and methylation accuracy, F1 scores, CNV concordance, and tumor–normal methylation contrast were evaluated. Performance was compared to whole-genome sequencing (WGS) controls and conventional bisulfite and enzymatic methylation sequencing methods.The chemistry demonstrated efficient reduction of modified cytosines, enabling high-confidence methylation calling with low false positive rates. Sequence complexity was maintained, allowing robust SNV and CNV detection from the same library. FFPE samples exhibited improved CpG coverage, fewer sequence artifacts, and improved library complexity compared to enzymatic methylation sequencing. cfDNA libraries generated from as little as 1 ng yielded reliable global methylation profiles suitable for tissue-of-origin analyses. The nondestructive nature of the chemistry preserved cfDNA fragments, opening the door to fragmentomics analyses. In tumor–normal comparisons, there was clear resolution of differentially methylated regions (DMRs), supporting more precise identification of tumor-specific epigenetic alterations. This enhanced positive-conversion chemistry enables unified genomic and epigenomic analysis from a single assay, making it highly suited for oncology research applications such as tumor classification, biomarker discovery, minimal residual disease (MRD) assessment, and liquid biopsy. Its gentle conditions, compatibility with degraded FFPE DNA and low-input cfDNA, and ability to capture both methylation and variants position it as a powerful tool for advancing translational research and precision oncology. Citation Format: Max Boeck, Jennifer Pavlica, Craig MARSHALL, Travis Sanders, Kristina Giorda, Martin Ranik, Eduard Casas, Thomas D. Harrison, Kailee Reed, Aaron Garnett, Doug Wendel, Brian Kudlow. Non-destructive methylation sequencing enables concurrent detection of genetic and epigenetic variation in FFPE and cfDNA samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1951.

The emerging field of precision oncology utilizes next-generation sequencing (NGS) to guide decisions based on individual genetic signatures. As demand for these applications grows, minimizing errors and biases introduced during library preparation and sequencing is critical. Advances in sequencing chemistries have improved base quality scores, shifting the focus to library preparation to improve data quality. PCR amplification is a key potential source of error during NGS library preparation; single-base substitutions, insertions, and deletions (indels) introduced by PCR in addition to PCR-induced biases in DNA length and GC content can distort data leading to decreased assay sensitivity and specificity. Improved library amplification, combined with recent innovations in sequencer technology, will increase the reliability of NGS assays and lead to better assay performance. Herein, we benchmark several commercial PCR amplification systems, including two solutions from Watchmaker Genomics: Equinox, and a re-engineered library amplification mix in development. We employed a UMI-powered NGS assay to quantify GC-uniformity, chimeric reads, base incorporation fidelity, and indel rates in microsatellite sequences. Libraries were sequenced on multiple platforms to leverage the ultra-high base quality scores of recent sequencing chemistries, resulting in low-noise measurements of polymerase fidelity. We also interrogate UMI family depth uniformity, a powerful proxy for template sequence bias and an informative indicator for error-correction efficacy. Furthermore, we assess amplification size bias by comparing the ability to amplify long templates in the presence of short templates. Compatibility with both SPRI and streptavidin beads was assessed. Finally, we test the ability to generate high terminal library yield, which is important for hybridization capture workflows. These results indicate that while some products may excel in one or two metrics, they dramatically underperform in others. Equinox, in particular the newly engineered variant, demonstrates robust performance across all relevant metrics, making it a leading solution for precise and unbiased library amplification. Giulia Corbet, Caitlyn Mendik, Lee French, Leo Karamanof, Meghan Oddy, Andrew Sutherland, Philip Benson, Josh Haimes, Ross Wadsworth, Travis Sanders, Craig Marshall, Kristina Giorda, Thomas Harrison, Eduard Casas, Brian Kudlow. A comprehensive benchmark study of multiple library amplification solutions for NGS [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2025; Part 1 (Regular Abstracts); 2025 Apr 25-30; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2025;85(8_Suppl_1):Abstract nr 6599.

Abstract Gene fusions due to chromosomal rearrangement, duplication, or deletion can be important drivers of cancer. Detection of gene fusions plays a valuable role in selecting targeted therapies. A widely used method for detecting fusion transcripts is targeted RNA sequencing, which focuses sequencing reads on known fusions but does not allow for the detection of novel fusions. Whole transcriptome analysis (WTA), where the entire transcriptome is interrogated, offers an opportunity to detect both known and novel fusions. Herein we examine the effect of sample quality, input mass, increased reverse transcriptase (RT) and insert size on fusion calling. Libraries are prepared using the Watchmaker RNA Library Preparation Kit with or without Polaris Depletion and, for target enrichment, the Twist Exome 2.0 panel prior to paired end 2 × 150 Illumina sequencing. First, control samples, specifically Seraseq Fusion RNA v4 high quality and FFPE samples, are used to determine if any of the parameters interrogated increase the number of breakpoint spanning reads both in a WTA and targeted sequencing context. Once established, the optimized workflow is applied to real FFPE samples. Targeted sequencing results in a higher number of breakpoint spanning reads, compared to WTA, when controlling for sequencing depth, resulting in a higher number of reads supporting each fusion call. Longer insert sizes, achieved by modified SPRI ratios, maintain library complexity and may enhance fusion calling. Taken together, these data demonstrate an optimized protocol for fusion calling from FFPE and demonstrate the utility of both WTA and targeted sequencing for fusion detection. Citation Format: Giulia Corbet, Josh Haimes, Travis Sanders, Martin Ranik, Thomas Harrison, Eduard Casas, Kailee Reed, Doug Wendel, Kristina Giorda, Ariele Hanek, Jen Pavlica, Brian Kudlow. Optimized RNA Seq library preparation for fusion calling from FFPE samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 321.

Abstract In the pursuit of advancing multi-cancer early detection, the exploration of cell-free DNA (cfDNA) through methylation profiling has emerged as a promising avenue. However, the realization of its full potential is impeded by limitations inherent in conventional methodologies, notably the 'gold standard' bisulfite conversion methods. These approaches pose challenges such as DNA destruction and the generation of low-complexity sequences resulting from the conversion of unmethylated cytosines to thymine. TET Assisted Pyridine-Borane Sequencing (TAPS) holds the potential as a non-destructive, rapid, and scalable alternative for base-level methylation analysis. In this context, we introduce TET Assisted Pyridine-Borane Sequencing version 1.0 (TAPS v1.0), showcasing advancements to the technology originally introduced by Schuster-Böckler, Song, and their collaborators. TAPS v1.0 transforms 5mC into dihydrouracil (DHU) through a series of oxidation and reduction reactions. Ultimately, DHU is recognized as dU during PCR resulting in C-to-T conversions at 5mC sites, leaving unmethylated cytosines, the majority of cytosines, unchanged. The nondestructive nature of 5mC conversion yields significantly higher library yield and complexity compared to BS-seq. By leveraging Watchmaker's proprietary polymerase evolution platform, a DNA polymerase tolerant to DHU was selected for library amplification post DHU conversion, enabling unimpeded detection of 5mC, particularly in densely packed CpG repeats. Coupled with an optimized End Repair and A-tailing chemistry preceding TAPS v1.0, clinically relevant cfDNA samples of low quality and/or quantity can be analyzed in methylation detection pipelines. To evaluate the performance of TAPS v1.0, human genomic DNA with control spike-ins featuring varying methylation statuses were employed. TAPS v1.0 exhibited significantly higher final library yield and greater complexity compared to BS-seq. Similar outcomes were observed in clinically relevant cfDNA samples. Citation Format: Josh Haimes, Craig Marshall, Travis Sanders, Sam Vogel, Martin Ranik, Eduard Casas, Doug Wendel, Leo Karamanof, Bjarne Faurholme, Abre de Beer, Lee French, Ross Wadsworth, Kristina Giordia, Eric van der Walt, Brian Kudlow. Advancing methylation analysis: TAPS v1.0: A fast, scalable, and non-destructive sequencing method [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 7025.

Abstract Formalin-fixed, paraffin-embedded (FFPE) samples are an invaluable resource in the oncology space, providing access to a vast library of archived diseased tissue samples paired with relevant donor information. Despite the broad utility of these samples, RNA extracted from FFPE tissue is typically difficult to process due to the presence of residual crosslinks and its degraded nature. Further, these samples often vary widely in performance, as the fixation process, block age, block storage, and extraction method can impart large impacts on resulting template quality. As a result, robust and reproducible RNA sequencing from FFPE-derived RNA remains a challenge with unpredictable and high failure rates. We evaluated four commercially available WTA solutions to determine which performed best with this challenging sample type with respect to library complexity, inter-input and intra-sample concordance, and overall reproducibility. Matched fresh frozen and FFPE liver samples were used such that the fresh frozen data set serves as a comparative truth set for the FFPE data. RNA-Seq libraries were prepared from each sample at both 10 ng and 100 ng inputs. The higher input serves as a control while the lower input reflects achievable inputs from very challenging FFPE samples. For 10 ng FFPE-derived RNA, the Watchmaker RNA Library Prep Kit with Polaris Depletion detects more unique genes than other chemistries and has a much higher percentage of genes that overlap with the fresh frozen control libraries. Additionally, differential expression analysis between 10 ng and 100 ng FFPE libraries demonstrates that the Watchmaker solution better preserves the gene expression profile at low inputs, as fewer differentially expressed genes are identified. The Watchmaker chemistry includes a number of features aimed at improving library complexity, including a novel FFPE decrosslinking step, a reverse transcriptase specifically engineered to improve conversion of RNA to cDNA, and fewer bead purification steps to prevent sample loss. The concordance between FF and FFPE samples at variable input masses observed when using the Watchmaker solution promotes a higher degree of confidence when making decisions based on lower quantity and quality samples and enables researchers to access more meaningful biological information. Citation Format: Jennifer Pavlica, Travis Sanders, David Gelagay, Kailee Reed, Thomas Harrison, Giulia Corbet, Josh Haimes, Ariele Hanek, Kristina Giorda, Brian Kudlow. A comparative analysis of library preparation technologies for RNA sequencing from FFPE samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2024; Part 1 (Regular Abstracts); 2024 Apr 5-10; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2024;84(6_Suppl):Abstract nr 2940.

Clinical oncology heavily relies on formalin-fixed, paraffin-embedded (FFPE) tissue samples for histology and molecular characterization. The chemical and physical modifications of nucleic acids introduced during fixation, storage and purification negatively impact molecular profiling and vary from sample-to-sample. Conventional sonication and ligation-based library preparation is considered the gold-standard approach for FFPE samples, but it is time-consuming and expensive. Importantly, these processes introduce artifacts that impact downstream analysis and interpretation (Haile et al., 2019). We have developed a novel fragmentation chemistry that virtually eliminates hairpin artifacts, achieving levels on par with non-FFPE control samples. Our fragmentation method is highly scalable, exhibits minimal sequence bias, and reduces cost and workflow inefficiencies associated with mechanical shearing. In this study, we developed a unified method for library preparation from FFPE samples which produces similar insert sizes across variable input mass and quality of FFPE samples. We carefully evaluated the performance of this method relative to a sonication-based approach employing the KAPA HyperPrep kit. Libraries were constructed from 50 to 200 ng of FFPE DNA, inputs typically used for NGS, ranging from low-to high-quality as assessed using a qPCR-based method and DNA integrity (DIN). Targeted sequencing was performed using a 37 kb custom oncology hybridization capture panel to investigate molecular complexity. Our workflow virtually eliminated hairpin artifacts that were present in up to 4.5% of reads in sonication-based libraries. Soft-clipping was also 3- to 7-fold lower in libraries prepared with the Watchmaker kit relative to sonicated DNA libraries, improving overall sequencing economy. Furthermore, the mean target coverage achieved with the Watchmaker kit was comparable to or higher than sonication libraries using the same input mass. Because input masses were normalized post-shearing, which typically results in 20-40% sample loss, coverage with our approach is significantly higher relative to sonication, if normalizing to pre-sheared DNA input. Watchmaker DNA Library Preparation with Fragmentation enables high-quality DNA library preparation from damaged FFPE samples, producing high target coverage, uniform insert size, and minimizing sequencing artifacts to improve sensitivity and specificity. This approach is highly scalable and automatable, enabling various oncology applications. Citation Format: Giulia Corbet, Philip Benson, Kailee Reed, Skyler Mishkin, Thomas Harrison, Kristin Scott, Zane Jaafar, Kristina Giorda, Josh Haimes, Martin Ranik, Brian Kudlow. A novel enzymatic library preparation workflow that dramatically reduces artifacts associated with damaged FFPE samples [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 215.

mRNA sequencing via poly(A) selection is a widely used and highly useful tool for molecular profiling but has limitations in the oncology space. The template damage of RNA extracted from formalin-fixed, paraffin-embedded (FFPE) material results in 5’ information loss. The hemoglobin mRNA content of blood-derived RNA necessitates a separate globin depletion step. Additionally, poly(A) enrichment results in a loss of a substantial proportion of long non-coding RNAs (lncRNAs), which are of growing interest as cancer biomarkers and therapeutic targets. Total RNA sequencing, where highly abundant, uninformative RNAs are specifically depleted, is more appropriate for these samples and applications, but traditional approaches are both labor- and time-intensive. To address this need, we developed a simple, rapid total RNA sequencing library preparation solution that improves data quality for the clinical and translational oncology space.We improved upon existing RNA depletion methods by modifying the chemistry of probe hybridization to enable a simplified workflow while also reducing unintended damage to non-targeted RNA. A reverse transcriptase was specifically engineered to improve the conversion of RNA to cDNA - a traditional bottleneck for RNA library preparation complexity. A novel FFPE decrosslinking step was implemented that may increase the amount of RNA available for downstream processing. Cleanup steps were minimized to reduce handling time and sample loss, and enzymatic steps were combined and shortened to simplify the overall workflow. Performance was assessed with RNA extracted from whole blood and multiple independent FFPE blocks of varying qualities as measured by DV200. Results highlight improved sequencing economy and increased gene detection sensitivity with low input and degraded samples. Highly concordant gene expression profiles were observed across a wide range of RNA input amounts, and data show excellent correlation between matched fresh frozen and FFPE samples. Lastly, gene fusions were accurately and confidently identified from an FFPE fusion control sample utilizing a whole transcriptome, non-targeted, approach.This work demonstrates the utility of a total RNA sequencing approach for oncology-relevant sample types, as well as the ability to improve on both workflow and data quality in a single library preparation solution. Citation Format: Travis Sanders, David Gelagay, Deelan Doolabh, Lee French, Jennifer Pavlica, Julie Walker, Clara Ross, Kailee Reed, Thomas Harrison, Ross Wadsworth, Eric van der Walt, Brian Kudlow. Improving whole transcriptome library preparation workflow and data quality for oncology-relevant samples and applications [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 242.

After a decade of NGS technology and workflow innovation, personalized medicine is starting to become a reality. DNA fragmentation is still a critical bottleneck during library preparation. Sonication methods have been the gold standard for consistent fragmentation and uniform GC coverage, but are associated with a high upfront investment, expensive consumables, and are prone to oxidative DNA damage. On the other hand, enzymatic fragmentation methods hold the potential to be scalable, automation-friendly workflows that minimize DNA damage. Yet, typical enzymatic fragmentation methods have been shown to introduce systematic hairpin artifacts, exhibit suboptimal uniformity, and have a narrow range of input mass compatibility. We have developed a novel enzymatic fragmentation-based library preparation technology which, together with our high-fidelity library amplification module, effectively overcomes many of the key limitations of related chemistries. Libraries generated using our workflow reduced chimeric reads and terminal hairpin artifacts 10-fold compared to other enzymatic methods, and reached comparable levels of mechanically sheared DNA controls. Furthermore, we observed consistently even coverage uniformity and low sequence-specific bias for human whole genome sequencing. Fragmentation was tested across a broad sample input range from 100 pg to 500 ng. Library insert sizes were highly tunable from 150 bp to 550 bp and consistent across the input titration. High quality libraries with minimal adapter dimer were prepared from as little as 1 picogram of gDNA by adjusting the post amplification clean up conditions. Taken together, this enzymatic fragmentation and library preparation workflow avoids library preparation artifacts that convolute variant calling, is highly scalable, and suitable for ultra-low input samples. Citation Format: Zane Jaafar, Josh Haimes, Thomas Harrison, Lindsay Peterkin, Martin Ranik, Kristin Scott, Brian A. Kudlow. A novel enzymatic fragmentation library preparation chemistry that greatly reduces sequencing artifacts [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2022; 2022 Apr 8-13. Philadelphia (PA): AACR; Cancer Res 2022;82(12_Suppl):Abstract nr 2288.