Refining the resolution of the yeast genotype-phenotype map using single-cell RNA-sequencing

Reviewer #1 (Public review):

In the revision of their paper, N'Guessan et al have improved the report of their study of expression QTL (eQTL) mapping in yeast using single cells. The authors make use of advances in single cell RNAseq (scRNAseq) in yeast to increase the efficiency with which this type of analysis can be undertaken. Building on prior research led by the senior author that entailed genotyping and fitness profiling of almost 100,000 cells derived from a cross between two yeast strains (BY and RM) they performed scRNAseq on a subset of ~5% (n = 4,489) individual cells. To address the sparsity of genotype data in the expression profiling they used a Hidden Markov Model (HMM) to infer genotypes and then identify the most likely known lineage genotype from the original dataset. To address the relationship between variance in fitness and gene expression the authors partition the variance to investigate the sources of variation. They then perform eQTL mapping and study the relationship between eQTL and fitness QTL identified in the earlier study.

This paper seeks to address the question of how quantitative trait variation and expression variation are related. scRNAseq represents an appealing approach to eQTL mapping as it is possible to simultaneously genotype individual cells and measure expression in the same cell. As eQTL mapping requires large sample sizes to identify statistical relationships, the use of scRNAseq is likely to dramatically increase the statistical power of such studies. However, there are several technical challenges associated with scRNAseq and the authors' study is focused on addressing those challenges. My main suggestion from my review of the revised version of the manuscript has been addressed in the revised figure 3. I agree with the authors that they have successfully demonstrated their stated goal of developing, and illustrating the benefit of, a one-pot scRNA-seq experiment and analysis for eQTL mapping.

Reviewer #2 (Public review):

This work describes the single-cell expression profiling of thousands of cells of recombinant genotypes from a model natural-variation system, a cross between two divergent yeast strains.

I appreciate the addition of lines 282-291, which now makes the authors' point about one advantage of the single-cell technique for eQTL mapping clearly: the authors don't need to normalize for culture-to-culture variation the way standard bulk methods do (e.g. in Albert et al., 2018 for the current yeast cross), and without this normalization, they can integrate analyses of expression with those of estimates of growth behaviors from the abundance of a genotype in the pool. The main question the manuscript addresses with the latter, in Figure 3, is how much variation in growth appears to have nothing to do with expression, for which the answer the authors given is 30%. I agree that this represents a novel finding. The caveats are (1) the particular point will perhaps only be interesting to a small slice of the eQTL research community; (2) the authors provide no statistical controls/error estimate or independent validation of the variance partitioning analysis in Figure 3, and (3) the authors don't seem to use the single-cell growth/fitness estimates for anything else, as Figure 4 uses loci mapped to growth from a previously published, standard culture-by-culture approach. It would be appropriate for the manuscript to mention these caveats.

I also think it is not appropriate for the manuscript to avoid a comparison between the current work and Boocock et al., which reports single-cell eQTL mapping in the same yeast system. I recommend a citation and statement of the similarities and differences between the papers.

I appreciate the new statement about the single-cell technique affording better power in eQTL mapping (lines 445-453).

Author response:

The following is the authors’ response to the previous reviews

Reviewer #1:

This paper seeks to address the question of how quantitative trait variation and expression variation are related. scRNAseq represents an appealing approach to eQTL mapping as it is possible to simultaneously genotype individual cells and measure expression in the same cell. As eQTL mapping requires large sample sizes to identify statistical relationships, the use of scRNAseq is likely to dramatically increase the statistical power of such studies. However, there are several technical challenges associated with scRNAseq and the authors' study is focused on addressing those challenges. Most of the points raised by my review of the initial version have been addressed. However, one point remains and one additional point should be considered. In this version the authors have introduced the use of data imputation using a published algorithm, DISCERN. This has greatly increased the variation explained by their model as presented in figure 3. However, it is possible that the explained variance is now an overestimation as a result of using the imputed expression data. I think that it would be appropriate to present figure 3 using the sparse data presented in the initial version of the paper and the newly presented imputed data so that the reader can draw their own conclusions about the interpretation.

We thank the reviewer for pointing this out and decided to present the results obtained from the sparse data in the main Figure 3 to avoid any overestimation. We also performed the variance partitioning at different sample sizes and used an optimized implementation of the GREML method to be able to handle high sample sizes instead of having to use a bootstrap estimate. As for the benefits of denoising the expression data, we illustrated it in the supplementary figure S6 so that people can draw their own conclusions about this imputation method. The imputation generally increases the contribution of the expressiongenotype interaction and decreases the residuals of the model by up to 8%.

Reviewer #1:

Given that the authors overcame many technical and analytical challenges in the course of this research, the study would be greatly strengthened through analysis of at least one, and ideally several, more conditions which would expand the conclusions that could be drawn from the study and demonstrate the power of using scRNAseq to efficiently quantify expression in different environments.

Our aim was to illustrate the benefit of one-pot scRNA-seq for eQTL mapping and the association of transcriptomic variation to trait variation. We think we have reached this goal with the current study. We understand that performing another scRNA-seq experiment in a new environment would help expand/validate our conclusions, but we think this would be a better fit for a future study.

Reviewer #2:

The authors now say the main take-home for their work is (1) they have established methods for linkage mapping with scRNA-seq and that these (2) "can help gain insights about the genotype-phenotype map at a broader scale." My opinion in this revision is much the same as it was in the first round: I agree that they have met the first goal, and the second theme has been so well explored by other literature that I'm not convinced the authors' results meet the bar for novelty and impact. To my mind, success for this manuscript would be to support the claim that the scRNA-seq approach helps "reveal hidden components of the yeast genotype-to-phenotype map." I'm not sure the authors have achieved this. I agree that the new Figure 3 is a nice addition-a result that apparently hasn't been reported elsewhere (30% of growth trait variation can't be explained by expression). The caveats are that this is a negative result that needs to be interpreted with caution; and that it would be useful for the authors to clarify whether the ability to do this calculation is a product of the scRNA-seq method per se or whether they could have used any bulk eQTL study for it. Beside this, I regret to say that I still find that the results in the revision recapitulate what the bulk eQTL literature has already found, especially for the authors' focal yeast cross: heritability, expression hotspots, the role of cis and transacting variation, etc.

We agree with the reviewer that this study does not reveal new modes of transcription regulation or phenomena that were not highlighted or hypothesized in the literature. To avoid confusion, we refrained from using the word “reveal” for such cases. However, we provide convincing evidence that one-pot scRNA-seq helps refining our understanding of genotype-phenotype map in two ways. First, the larger scalability of this approach allowed us to find a median number of eQTL per gene that is ~4 times higher than the largest bulk-eQTL mapping in the same genetic background. For 60% of these genes, i.e. the ones with higher expression heritability in our dataset, the ability to explain their transcriptomic variation from SNPs increased by ~16% on average, which is substantial. This gain in power can thus improve our understanding of the gene network by highlighting new downstream effects of mutations or transcriptome variation. Second, by performing one-pot eQTL as opposed to large-scale bulk eQTL, thousands of transcriptomes can be collected simultaneously without having to use batching strategies. This enables the association between phenotype, genotype and expression variation, which we show in figure 3 through variance partitioning. While it is possible that the growth trait variation not being fully explained by expression could be an artifact of scRNA-seq, we do not believe this is the case because most transcriptional variation is explained by genotype (~76%).

Furthermore, we show that by having to control expression for growth, by missing some hotspots of regulation and by missing multiple eQTL for each gene, previous bulk-eQTL analysis could not replicate the significant association between eQTL hotspots and QTL hotspot, which this study highlights. Thus, we agree in general that many of the insights about transcriptional regulation have been obtained through ‘brute-force’, bulk RNA-seq, which fundamentally can reach tens of thousands of transcriptomes as well, but we believe the one-pot scRNA-seq approach is much easier and expedient once genotyping the single-cells and other challenges regarding denoising and low coverage have been solved (which we believe we did). There is indeed another reviewed preprint [Boocock et al, eLife] that has used similar approaches as our study since the publication of our manuscript (in October 2023).

Likewise, when in the first round of review I recommended that the authors repeat their analyses on previous bulk RNA-seq data from Albert et al., my point was to lead the authors to a means to provide rigorous, compelling justification for the scRNA-seq approach. The response to reviewers and the text (starting on line 413) says the comparison in its current form doesn't serve this purpose because Albert et al. studied fewer segregants. Wouldn't down-sampling the current data set allow a fair comparison? Again, to my mind what the current manuscript needs is concrete evidence that the scRNA-seq method per se affords truly better insights relative to what has come before.

We agree that down-sampling the current dataset would allow for a fair comparison. Thus, we illustrate the results of the variance partitioning at different sample sizes. While the total variance explained is similar, the contribution of the genotype-expression interaction increases with sample size, highlighting the increase in the confidence of the associations between expression and genotype that contributed to trait variation. We also showed that a lot of important low-effect sizes eQTL are missing at a sample size of 1000 compared to a sample size 4000. Indeed, by increasing the scale of eQTL mapping by ~4, about 60% of genes have increased heritability and this increase is due to eQTLs that cumulatively explain more than 15% of transcript level variation.

I also recommend that the authors take care to improve the main text for readability and professionalism. It would benefit from further structural revision throughout (especially in the figure captions) to allow high-impact conclusions to be highlighted and low-impact material to be eliminated. Figure 4 and the results text sections from line 319 onward could be edited for concision or perhaps moved to supplementary if they obscure the authors' case for the scRNA-seq approach. The text could also benefit from copy editing (e.g. three clauses starting with "while" in the paragraph starting on line 456; "od ratio" on line 415). I appreciate the authors' work on the discussion, including posing big picture questions for the field (lines 426-429), but I don't see how they have anything to do with the current scRNA-seq method.

We thank the reviewer for their suggestions for improving the readability of the text. We edited some of the figure captions and result section titles to better highlight the main results. However, we do not think that the last result section obscures our findings but rather supports the fact that scRNA-seq refines our understanding of the GPM. Indeed, we discovered many new eQTLs that are related to both expression and trait variation, highlighting the potential for understanding the downstream effects of mutations on the gene network and on trait variation through multiple trans-regulation paths.