Comprehensive benchmark and architectural analysis of deep learning models for nanopore sequencing basecalling.

BACKGROUND: Nanopore-based DNA sequencing relies on basecalling the electric current signal. Basecalling requires neural networks to achieve competitive accuracies. To improve sequencing accuracy further, new models are continuously proposed with new architectures. However, benchmarking is currently not standardized, and evaluation metrics and datasets used are defined on a per publication basis, impeding progress in the field. This makes it impossible to distinguish data from model driven improvements. RESULTS: To standardize the process of benchmarking, we unified existing benchmarking datasets and defined a rigorous set of evaluation metrics. We benchmarked the latest seven basecaller models by recreating and analyzing their neural network architectures. Our results show that overall Bonito's architecture is the best for basecalling. We find, however, that species bias in training can have a large impact on performance. Our comprehensive evaluation of 90 novel architectures demonstrates that different models excel at reducing different types of errors and using recurrent neural networks (long short-term memory) and a conditional random field decoder are the main drivers of high performing models. CONCLUSIONS: We believe that our work can facilitate the benchmarking of new basecaller tools and that the community can further expand on this work.

p53 Forms Redox-Dependent Protein-Protein Interactions through Cysteine 277.

Reversible cysteine oxidation plays an essential role in redox signaling by reversibly altering protein structure and function. Cysteine oxidation may lead to intra- and intermolecular disulfide formation, and the latter can drastically stabilize protein-protein interactions in a more oxidizing milieu. The activity of the tumor suppressor p53 is regulated at multiple levels, including various post-translational modification (PTM) and protein-protein interactions. In the past few decades, p53 has been shown to be a redox-sensitive protein, and undergoes reversible cysteine oxidation both in vitro and in vivo. It is not clear, however, whether p53 also forms intermolecular disulfides with interacting proteins and whether these redox-dependent interactions contribute to the regulation of p53. In the present study, by combining (co-)immunoprecipitation, quantitative mass spectrometry and Western blot we found that p53 forms disulfide-dependent interactions with several proteins under oxidizing conditions. Cysteine 277 is required for most of the disulfide-dependent interactions of p53, including those with 14-3-3θ and 53BP1. These interaction partners may play a role in fine-tuning p53 activity under oxidizing conditions.

The Human 2-Cys Peroxiredoxins form Widespread, Cysteine-Dependent- and Isoform-Specific Protein-Protein Interactions.

Redox signaling is controlled by the reversible oxidation of cysteine thiols, a post-translational modification triggered by H2O2 acting as a second messenger. However, H2O2 actually reacts poorly with most cysteine thiols and it is not clear how H2O2 discriminates between cysteines to trigger appropriate signaling cascades in the presence of dedicated H2O2 scavengers like peroxiredoxins (PRDXs). It was recently suggested that peroxiredoxins act as peroxidases and facilitate H2O2-dependent oxidation of redox-regulated proteins via disulfide exchange reactions. It is unknown how the peroxiredoxin-based relay model achieves the selective substrate targeting required for adequate cellular signaling. Using a systematic mass-spectrometry-based approach to identify cysteine-dependent interactors of the five human 2-Cys peroxiredoxins, we show that all five human 2-Cys peroxiredoxins can form disulfide-dependent heterodimers with a large set of proteins. Each isoform displays a preference for a subset of disulfide-dependent binding partners, and we explore isoform-specific properties that might underlie this preference. We provide evidence that peroxiredoxin-based redox relays can proceed via two distinct molecular mechanisms. Altogether, our results support the theory that peroxiredoxins could play a role in providing not only reactivity but also selectivity in the transduction of peroxide signals to generate complex cellular signaling responses.

Mitochondria-Derived H2O2 Promotes Symmetry Breaking of the C. elegans Zygote.

Symmetry breaking is an essential step in cell differentiation and early embryonic development. However, the molecular cues that trigger symmetry breaking remain largely unknown. Here, we show that mitochondrial H2O2 acts as a symmetry-breaking cue in the C. elegans zygote. We find that symmetry breaking is marked by a local H2O2 increase and coincides with a relocation of mitochondria to the cell cortex. Lowering endogenous H2O2 levels delays the onset of symmetry breaking, while artificially targeting mitochondria to the cellular cortex using a light-induced heterodimerization technique is sufficient to initiate symmetry breaking in a H2O2-dependent manner. In wild-type development, both sperm and maternal mitochondria contribute to symmetry breaking. Our findings reveal that mitochondrial H2O2-signaling promotes the onset of polarization, a fundamental process in development and cell differentiation, and this is achieved by both mitochondrial redistribution and differential H2O2-production.

Application of Recombination -Induced Tag Exchange (RITE) to study histone dynamics in human cells.

In eukaryotes, nucleosomes form a barrier to DNA templated reactions and must be dynamically disrupted to provide access to the genome. During nucleosome (re)assembly, histones can be replaced by new histones, erasing post-translational modifications. Measuring histone turnover in mammalian cells has mostly relied on inducible overexpression of histones, which may influence and distort natural histone deposition rates. We have previously used recombination-induced tag exchange (RITE) to study histone dynamics in budding yeast. RITE is a method to follow protein turnover by genetic switching of epitope tags using Cre recombinase and does not rely on inducible overexpression. Here, we applied RITE to study the dynamics of the replication-independent histone variant H3.3 in human cells. Epitope tag-switching could be readily detected upon induction of Cre-recombinase, enabling the monitoring old and new H3.3 in the same pool of cells. However, the rate of tag-switching was lower than in yeast cells. Analysis of histone H3.3 incorporation by chromatin immunoprecipitation did not recapitulate previously reported aspects of H3.3 dynamics such as high turnover rates in active promoters and enhancers. We hypothesize that asynchronous Cre-mediated DNA recombination in the cell population leads to a low time resolution of the H3.3-RITE system in human cells. We conclude that RITE enables the detection of old and new proteins in human cells and that the time-scale of tag-switching prevents the capture of high turnover events in a population of cells. Instead, RITE might be more suited for tracking long-lived histone proteins in human cells.