Does this suggest that immersion in any liquid would induce degradation? I would like to understand if the biocomposites start to degrade in water, or even in high-humidity environments without suspension in liquids. I might be missing something about what these lipases need for activity, though.

I'm curious about applications and materials where this would matter. Are there plastics that don't need as much elongation?

What could be contributing to reduced catalytic activity? It's interesting that activity is renewed after a vegetative cycle - do displayed lipases degrade over cycles?

I'm not very familiar with analyzing GPC traces. There are red and blue boxes highlighting peaks in the data around elutions times of 3.5-4min, but I don't see descriptions of those highlight boxes anywhere. Are these peaks important? I assume they are because they decrease over incubation time.

This is a fascinating preprint, thank you! Creating a more resilient way to break down plastics--especially by creating biocomposite materials--seems like a broadly useful breakthrough.

I'm curious about real-world application of this technology. Do you see either of the two processes described being more useful for manufacturers/processing facilities? Will it be easier to incorporate a spore suspension step, or is starting from a biocomposite material simpler?

Thank you for providing a wonderful tool and some really interesting insights from Drosophila! I noticed that the Github repo contains instructions for installing on multiple systems, and I'm wondering if providing a Nextflow or Snakemake-based approach would be useful for general users. I could see this being especially useful for the multi-sample input (both the many reads vs. single assembly and many reads vs. many assemblies), so users could provide a single input folder with minimal nested for looping.

This is the first use case I thought of while reading the manuscript - finding recent HGT events beyond coding regions could be an extremely powerful use of this tool. I'd love to see an example of this!

Would you be able to provide insight on what "high quality" means in this case? I assume the bar to confidently call small gaps against fragmented reads is much higher than in some other cases, so some guidance on what level of quality is warranted here would be great.

Oh, I see this section guides users through finding candidates. I wonder if this could be incorporated into the earlier section (or point to this section), since it's helpful.

I'm a little confused how I'd find the most promising sequence from looking at Figure 2D without knowing KoRV was present for that particular sequence. Are the three measures mentioned the best for gauging this, and is there a weighting to apply to each - for example, is low coverage bias more important than substantial length, or should they be considered equally? Also, what defines "substantial length" in the context of a TE?

Could you comment more on the impact of degraded/endogenized transposable element sequences on the predictive accuracy of GenomeDelta? I imagine that remnants generated over decades or centuries would significantly obfuscate the results for that particular element, but I'm not sure if that is accurate?

This is a wonderful paper with clear results and great explanations. I'm super curious about the cargo as well. Is it common for single genes (or two in this case) to have similar effects on male AND female fertility, or are differences between males and females more common? It seems striking to me that the knockout of Pnma1 and Pnma4 dramatically reduced fertility across the board. I wonder if looking at other genes that have similar impacts in both sexes would help to understand what the cargo is potentially interacting with.

Is there a potential spatial component to this? If capsids are used to transmit signals, then would identifying the location of the PNMA1 and PNMA4-expressing populations in relation to nearby cell types give you potentially useful information?

You found that PNMA1 and PNMA4 expression reduces with age - would you then expect capsid production also decline with age?

Do you have a sense of how difficult/how efficient this procedure is for non-plant tissue samples?

I would also be curious about other age-related biomarkers beyond memory and fat accumulation!

Will the code to run these analyses be available on Github? It would be helpful to follow along with the many steps used across these analyses.

I was a little confused on the workflow in this section.

I really, really appreciate the use of public data to ask a unique question!

I could totally see follow-up experimental work examining fecundity and offspring survival rates in yolkless vs. yolk groups of endoparasitoids!

I wonder whether these other genes may be lost in the transition of egg yolk loading to harnessing another organism’s nutrients for development, rather than after? I think I may just not be super clear on how this is able to tell that these changes happened after the loss of Vg.

This dataset seems like it has the capacity to look at genes beyond Vg and VgR - I would be very curious to see if there are any other genes, especially ones associated with observable phenotypes (e.g., egg size, developmental timeline, features, etc.), that can be interrogated the same way.

This is a really interesting finding. Do you think there are host traits that may be associated with Vg loss? Maybe certain hosts have "richer" hemolymph, which makes it more likely that a parasitoid will lose its own yolk? Or is it more likely that this is purely a genome rearrangement phenomenon?

Why are Vg genes so commonly found at genome rearrangement breakpoints? I don’t know very much about where these rearrangement breakpoints fall, but are there unique genomic factors surrounding these genes that make them more prone to loss?

On the note of genome assemblies, I really appreciate your inclusion of the metadata in Supplementary Table 11. Out of curiosity, how much do you think the quality of genome assemblies impacts your results more broadly?

Does this finding then support the previous result that A. japonica lost its yolk?

Are there any interesting patterns with Vg copy number beyond presence/absence? It would be interesting to know more about the multiple copies present in the non-parasitoid hymenopterans. I'm especially curious about multiple copies in parasitoid species - from a quick glance, the only loss events seem to be in species related to those with only one Vg copy.

What a unique study on a wonderful set of organisms using available data! Thank you for doing this research.