DNAscent v2: detecting replication forks in nanopore sequencing data with deep learning

Abstract

Background Measuring DNA replication dynamics with high throughput and single-molecule resolution is critical for understanding both the basic biology behind how cells replicate their DNA and how DNA replication can be used as a therapeutic target for diseases like cancer. In recent years, the detection of base analogues in Oxford Nanopore Technologies (ONT) sequencing reads has become a promising new method to supersede existing single-molecule methods such as DNA fibre analysis: ONT sequencing yields long reads with high throughput, and sequenced molecules can be mapped to the genome using standard sequence alignment software. Results This paper introduces DNAscent v2, software that uses a residual neural network to achieve fast, accurate detection of the thymidine analogue BrdU with single-nucleotide resolution. DNAscent v2 also comes equipped with an autoencoder that interprets the pattern of BrdU incorporation on each ONT-sequenced molecule into replication fork direction to call the location of replication origins termination sites. DNAscent v2 surpasses previous versions of DNAscent in BrdU calling accuracy, origin calling accuracy, speed, and versatility across different experimental protocols. Unlike NanoMod, DNAscent v2 positively identifies BrdU without the need for sequencing unmodified DNA. Unlike RepNano, DNAscent v2 calls BrdU with single-nucleotide resolution and detects more origins than RepNano from the same sequencing data. DNAscent v2 is open-source and available at https://github.com/MBoemo/DNAscent. Conclusions This paper shows that DNAscent v2 is the new state-of-the-art in the high-throughput, single-molecule detection of replication fork dynamics. These improvements in DNAscent v2 mark an important step towards measuring DNA replication dynamics in large genomes with single-molecule resolution. Looking forward, the increase in accuracy in single-nucleotide resolution BrdU calls will also allow DNAscent v2 to branch out into other areas of genome stability research, particularly the detection of DNA repair.

Genome-wide mapping of individual replication fork velocities using nanopore sequencing

“…1a ). RepNano 13 and other published BrdU basecallers, DNAscent 15 and DNAscent v2 16 , were also assessed for comparison. Megalodon exhibited the lowest background signal without BrdU and most closely paralleled the perfect correlation line, despite a slight tendency to underestimate BrdU content (Supplementary Fig.…”

“…For each of the 11 genomic DNA samples with different BrdU substitution rates described above, a set of 8000 nanopore reads were basecalled either with Megalodon or with DNAscent v1 15 , DNAscent v2 16 , RepNano 13 transition matrices (RepNano_TM) and RepNano convolutional neural network (RepNano_CNN). RepNano_TM and RepNano_CNN were used as in ref.…”

Replication origin mapping in the malaria parasite Plasmodium falciparum

“…The thymidine analogues would thus mark DNA replication forks that had been active within the past 15 minutes, allowing us to confidently identify replication fork directionality and speed, and to infer the placement of active replication origins (Figure 4a and b). Sequencing data was analysed using DNAscent software v3.0.2, an overhauled version of its predecessor, which originally detected BrdU nucleotide incorporation into replicating DNA and identified fork direction and speed via the changing gradient of BrdU incorporation in the nascent strand (Boemo, 2021). To develop the most recent version, DNAscent was trained to distinguish between EdU and BrdU on the same molecule in order to more accurately and reliably identify fork directionality and origins.…”

“…B) Nine single molecules sequenced on the Oxford nanopore MinION platform that mapped to P. falciparum chromosome 1 and were analysed with DNAscent v3.02. Similar to the visualization in (Boemo, 2021), each molecule is represented as two bedgraph tracks showing the probability of BrdU (red) and EdU (blue) called by DNAscent v3.0.2 at each thymidine position in the read. C) Visualisation showing fired origins called by DNAscent at 30 and 36 hpi on chromosome 2 (red) with ORC1 ChIP peaks (positive peaks in blue, negative peaks in grey) for comparison.…”

“…In model systems this has been achieved in various ways, including chromatin immunoprecipitation (ChIP) for proteins such as ORC1, and mapping of nascent DNA synthesis by diverse means, including incorporation of modified nucleotides followed by their purification and mapping (Eaton et al ., 2011). We have combined ORC1-ChIP with a novel application of the recently published nanopore methods (Boemo, 2021; Georgieva et al, 2020; Hennion et al, 2020; Muller et al, 2019; Theulot et al, 2022), which exploits Oxford nanopore long-read sequencing to simultaneously detect and map modified nucleotides in nascent DNA strands. For the first time in any system, we have developed this into a powerful two-nucleotide protocol with new DNAscent software that distinguishes sequential pulses of the thymidine analogues bromo-deoxyuridine (BrdU) or ethyl-deoxyuridine (EdU) in single molecules.…”