The standard plots of modern bioinformatics, from genome browsers to survival curves, read through one example: how losing the tumor suppressor TP53 can leave its target p21 switched off.
Biology now measures thousands of cells, millions of DNA bases and whole patient cohorts at once, and bioinformatics has built a set of standard plots for reading data of that size.
The series follows the tumor suppressor TP53 and its target CDKN1A (p21). The first essay looks at them through five kinds of data at once. The rest take one kind each: coordinates and reads on the genome, genome-wide rearrangements, where p53 binds DNA, which genes change expression, how genes and patients cluster, how cells are mapped in two dimensions, and what the evidence says about patients. The genome is real wherever a figure is about the genome: chromosome lengths and bands, gene models, CpG islands, conservation, human-mouse synteny, the p53 motif and published p53 binding sites come from UCSC, Ensembl, JASPAR and ReMap, on GRCh38, and the TP53 mutations in Part 1 are the TCGA PanCancer Atlas as served by cBioPortal. Reads, expression values, cells and patients are simulated in the page, and each caption says which is which.
Read in order, the essays build one case. Part 1 finds four routes to a silent CDKN1A in a simulated cohort: TP53 mutation, MDM2 amplification, methylation of the CDKN1A promoter in tumors whose TP53 is intact, and none of these. Part 2 opens the real CDKN1A locus, and Part 4 asks whether the R175H mutant still binds its two p53 sites. Part 5 knocks out TP53 and finds CDKN1A among the genes that fall, and the rest of p53's targets falling together as a set though few pass a gene-by-gene test. Part 6 recovers the routes as clusters from the molecular data alone, and misses the partly methylated tumors. Part 8 asks whether any of it matters for survival: low CDKN1A does, and once CDKN1A is known the clusters add nothing. Parts 3 and 7 step off the thread, to the whole genome and to single cells.
One rule holds in every essay. Brick red means higher, gained or more: an up-regulated gene, an amplification, more methylation, a higher hazard. Blue means lower, lost or less. Teal marks whatever is in focus, usually CDKN1A or the site, cell or patient you picked. A mutation is not a direction, so mutant groups are drawn in teal or in a subtype color, with the wildtype reference in gray.