Benchmarking microbial growth rate predictions from metagenomes

Abstract

Growth rates are central to understanding microbial interactions and community dynamics. Metagenomic growth estimators have been developed, specifically codon usage bias (CUB) for maximum growth rates and "peak-to-trough ratio" (PTR) for in situ rates. Both were originally tested with pure cultures, but natural populations are more heterogeneous, especially in individual cell histories pertinent to PTR. To test these methods, we compared predictors with observed growth rates of freshly collected marine prokaryotes in unamended seawater. We prefiltered and diluted samples to remove grazers and greatly reduce virus infection, so net growth approximated gross growth. We sampled over 44 h for abundances and metagenomes, generating 101 metagenome-assembled genomes (MAGs), including Actinobacteria, Verrucomicrobia, SAR406, MGII archaea, etc. We tracked each MAG population by cell-abundance-normalized read recruitment, finding growth rates of 0 to 5.99 per day, the first reported rates for several groups, and used these rates as benchmarks. PTR, calculated by three methods, rarely correlated to growth (r ~-0.26-0.08), except for rapidly growing γ-Proteobacteria (r ~0.63-0.92), while CUB correlated moderately well to observed maximum growth rates (r = 0.57). This suggests that current PTR approaches poorly predict actual growth of most marine bacterial populations, but maximum growth rates can be approximated from genomic characteristics.

Fig. 1. Phylogenomic tree of metagenomic-assembled genomes from our experiments.

Numbers next to branches are bootstrap values (only bootstrap values ≥50 are shown here).

Fig. 2. Phylogenomic tree with associated heatmap showing the highest observed growth for each taxon in each experiment and replicate.

The detailed taxonomy information and observed growth rates can be found in Supplementary Table 1.

Fig. 3. Box plot of the highest observed growth rate for each MAG, grouped by taxonomic family.

The boxes are drawn from the 25th to the 75th quantiles and the center line of each box is the median. Whiskers indicate the smallest and largest values for each family.

Fig. 4. The highest observed growth rates, as measured over three time points, plotted against predicted maximum growth rates from codon usage bias predictor (growthpred).

Solid line is x = y, so MAGs with symbols above the line grew more slowly than its predicted maximum, and below the line had faster growth than its predicted maximum. Note that when growth exceeded the predicted maximum, it was usually not by much. The dashed line is the linear regression between observed growth and predicted maximum growth, with the correlation, equation, and p value shown.

Fig. 5. Observed growth rate, measured over three consecutive time points, plotted against peak-to-trough ratio indices at the middle time point.

The PTR indices are a iRep, b GRiD, and c DEMIC. Lines are linear regressions for each taxon with more than two observations of both growth rates and PTR indices. Slopes should be positive and significant if the PTR index reflects growth. The similar extent of positive, flat, and negative slopes illustrates the poor general relationship between PTR and observed growth rates, with the exception of a few fast-growing Gammaproteobacteria (see text). Underlying data and regression statistics are in Supplementary Table 2.