In general, abstraction is a process to conceptualize real-world observations (such as biological phenomenon) into generic models (symbolic models). For computational analysis, the abstract models is also the starting point.
Newick representation of trees is an abstraction of the graphs.
Pipeline analysis in genome analysis is an example of this.
PCR is one of the best example of "recursion" application in biology, though somewhat unknowingly.
Search space is a fundamental concept in quantitative reasoning. Google made its name by efficiently narrowing down the internet searching space.
In bioinformatics, search space is illustrated in dot plot and dynamic programming approach for sequence alignment. NCBI BLAST is a fast program because it only searches in selected regions in the search space. With properly designed methods, we can greatly reduce the number of steps needs to find a solution in a search space.
In biology, it is also a wonder how the very few copies of transcription factors can effectively find its binding motifs along chromosomes.
Be able to visualize quantitative results and interpret them.
Phylogenetic trees is probably a classical example.
Permutation are often used to evaluate the significance of biological observations.
Break down of a large problem into manageable ones. ClustalW use pairwise-approach to progressively build a multiple alignment, guided by a tree that describes the combination of pairs at each step.
- Tree thinking (Phylogeny)
Phylogeny is the basic concept in biology.
- Modularization
- Normalization
To compare samples generated by different experiments, we need to put them on the same ground. This is why microarray data and sequence profiles need to be normalized.
- Optimization
- Computational complexity
- Greedy method
ClustalW is a greedy method for multiple sequence alignment because it looks for the best pairs (local optimal solutions) in each step.
Gene/protein interaction networks
In the Smith-Waterman algorithm for pairwise sequence alignment, the best direction is chosen at each step from 3 possible directions. The high-score-segment is constructed by going backward along the marked directions.
493 visits.