Scientific methods have the dual role of innovation as well as to root out error. Modern science does not just rely on either theory or experimentation but a vast majority of current contributions are based on integration of the plethora of data that already exists, data that has been gathered by in vivo/vitro and also by silico methods.
With the ever increasing curiosity of understanding biological systems at various degrees of details and depths, data integration is also being used in a number of challenging ways, it could be to gather information from complementary aspects of a system, like, clinical, genetic, genomic, proteomic data. It might also be using various experimental scenarios for a single aspect, more like in a meta-analytic framework.
These methods generate a myriad of hypotheses which provide interesting directions of further research among other outcomes.
However, the main achievement of a method is always how closely and how correctly we captured the underlying system of interest. Which brings up critical questions like accuracy, robustness, reproducibility. Accuracy and robustness are theoretical achievements which hold true if the underlying assumptions are true. Reproducibility also depends a lot on assumptions however it is gets tested in real life.
Reproducibility itself has many facets to consider, a few prominent ones would be data/empirical reproducibility, statistical/methodological reproducibility, computational reproducibility.
Our objective here is to showcase different aspects of data integration that bring out the strengths and weaknesses of the underlying methods when judged by its validity in future and that can be measured today. There's a plethora of activities from known disciplines of computer science, statistics, mathematics, bio computing, and also activities in less known fields like artificial intelligence that we hope to cover under this Research Topic.
Scientific methods have the dual role of innovation as well as to root out error. Modern science does not just rely on either theory or experimentation but a vast majority of current contributions are based on integration of the plethora of data that already exists, data that has been gathered by in vivo/vitro and also by silico methods.
With the ever increasing curiosity of understanding biological systems at various degrees of details and depths, data integration is also being used in a number of challenging ways, it could be to gather information from complementary aspects of a system, like, clinical, genetic, genomic, proteomic data. It might also be using various experimental scenarios for a single aspect, more like in a meta-analytic framework.
These methods generate a myriad of hypotheses which provide interesting directions of further research among other outcomes.
However, the main achievement of a method is always how closely and how correctly we captured the underlying system of interest. Which brings up critical questions like accuracy, robustness, reproducibility. Accuracy and robustness are theoretical achievements which hold true if the underlying assumptions are true. Reproducibility also depends a lot on assumptions however it is gets tested in real life.
Reproducibility itself has many facets to consider, a few prominent ones would be data/empirical reproducibility, statistical/methodological reproducibility, computational reproducibility.
Our objective here is to showcase different aspects of data integration that bring out the strengths and weaknesses of the underlying methods when judged by its validity in future and that can be measured today. There's a plethora of activities from known disciplines of computer science, statistics, mathematics, bio computing, and also activities in less known fields like artificial intelligence that we hope to cover under this Research Topic.
