Treasures from trash in cancer research

Abstract

Introduction: Cancer research has significantly improved in recent years, primarily due to next-generation sequencing (NGS) technology. Consequently, an enormous amount of genomic and transcriptomic data has been generated. In most cases, the data needed for research goals are used, and unwanted reads are discarded. However, these eliminated data contain relevant information. Aiming to test this hypothesis, genomic and transcriptomic data were acquired from public datasets. Materials and Methods: Metagenomic tools were used to explore genomic cancer data; additional annotations were used to explore differentially expressed ncRNAs from miRNA experiments, and variants in adjacent to tumor samples from RNA-seq experiments were also investigated. Results: In all analyses, new data were obtained: from DNA-seq data, microbiome taxonomies were characterized with a similar performance of dedicated metagenomic research; from miRNA-seq data, additional differentially expressed sncRNAs were found; and in tumor and adjacent to tumor tissue data, somatic variants were found. Conclusions: These findings indicate that unexplored data from NGS experiments could help elucidate carcinogenesis and discover putative biomarkers with clinical applications. Further investigations should be considered for experimental design, providing opportunities to optimize data, saving time and resources while granting access to multiple genomic perspectives from the same sample and experimental run.

Toward a Predictive Understanding of Cyanobacterial Harmful Algal Blooms through AI Integration of Physical, Chemical, and Biological Data

“…We recognize that while cyanoHABs are dominated by cyanobacteria, there may be other actors, notably viruses, other bacteria, and eukaryotic microbes present in the water samples. Existing sequence analysis tools, e.g., read aligners such as Bowtie2 and de novo assemblers such as SPAdes, will need additional tuning and computation scaling to detect the presence and abundance of fragment-based features that are routinely filtered out in conventional meta-‘omics pipelines. , …”

“…Existing sequence analysis tools, e.g., read aligners such as Bowtie2 107 and de novo assemblers such as SPAdes, 108 will need additional tuning and computation scaling to detect the presence and abundance of fragment-based features that are routinely filtered out in conventional meta-‘omics pipelines. 109 , 110 …”

“…Existing sequence analysis tools, e.g., read aligners such as Bowtie2 107 and de novo assemblers such as SPAdes, 108 will need additional tuning and computation scaling to detect the presence and abundance of fragment-based features that are routinely filtered out in conventional meta-'omics pipelines. 109,110 Figure 4 outlines a suggested workflow to build predictive and interpretative ML models for HABs that may be used by domain experts and decision-makers.…”