Computational surprisal analysis speeds-up genomic characterization of cancer processes

Surprisal analysis of genome-wide transcript profiling identifies differentially expressed genes and pathways associated with four growth conditions in the microalga Chlamydomonas

The usual cultivation mode of the green microalga Chlamydomonas is liquid medium and light. However, the microalga can also be grown on agar plates and in darkness. Our aim is to analyze and compare gene expression of cells cultivated in these different conditions. For that purpose, RNA-seq data are obtained from Chlamydomonas samples of two different labs grown in four environmental conditions (agar@light, agar@dark, liquid@light, liquid@dark). The RNA seq data are analyzed by surprisal analysis, which allows the simultaneous meta-analysis of all the samples. First we identify a balance state, which defines a state where the expression levels are similar in all the samples irrespectively of their growth conditions, or lab origin. In addition our analysis identifies additional constraints needed to quantify the deviation with respect to the balance state. The first constraint differentiates the agar samples versus the liquid ones; the second constraint the dark samples versus the light ones. The two constraints are almost of equal importance. Pathways involved in stress responses are found in the agar phenotype while the liquid phenotype comprises ATP and NADH production pathways. Remodeling of membrane is suggested in the dark phenotype while photosynthetic pathways characterize the light phenotype. The same trends are also present when performing purely statistical analysis such as K-means clustering and differentially expressed genes.

Metabolomic analysis of the green microalga Chlamydomonas reinhardtii cultivated under day/night conditions

Biomass composition of Chlamydomonas reinhardtii was studied during two consecutive cycles of 12h light/12h dark. As in our experimental conditions the two synchronized divisions were separated by 20h, it was possible to show that accumulation of dry weight, proteins, chlorophyll and fatty acids mainly depends on cell division, whereas starch accumulation depends on a circadian rhythm as reported previously. Our metabolomics analyses also revealed that accumulation of five (Ser, Val, Leu, Ile and Thr) of the nine free amino acids detected displayed rhythmicity, depending on cell division while Glu was 20-50 times more abundant than the other ones probably because this free amino acid serves not only for protein synthesis but also for biosynthesis of nitrogen compounds. In addition, we performed a thermodynamic-motivated theoretical approach known as 'surprisal analysis'. The results from this analysis showed that cells were close to a steady state all along the 48h of the experiment. In addition, calculation of free energy of cellular metabolites showed that the transition point, i.e. the state which immediately precedes cell division, corresponds to the most unstable stage of the cell cycle and that division is identified as the greatest drop in the free energy of metabolites.