Add ability to highlight points to scatterplot

[asizemore]

Resolves #1322

This was a little funky because plotly already allows for highlighting - we just have to set the color appropriately in the trace! So this PR is more of a hijack of our current processing of scatterplot data to recolor the points.

A few considerations:

• Should the backend just break up the data into highlight and non-highlight for us? It could, and would basically add another "overlay"-type column, but the issue is when we highlight a scatterplot with anything that we have calculated. Smoothed means, best fit line, etc. Breaking up the data for highlgihting would move these points to a place that doesn't correspond with their place in a calculation. We don't want a highlight-smoothed mean and a non-highlight-smoothed mean, we would just want one smoothed mean with a few points highlighted. But we still want to know those highlighted points contributed to the calculation of the smoothed mean. This gets extra complicated with overlays and facets.
• On the frontend, should we recolor in place or make a new trace? Make a new trace so that there's an extra line in the legend that we can click on and off
• Where should the highlighting go on the frontend? Ideally in Scatterplot but that means we have to traverse the entire dataset again. We already traverse the dataset in processInputData, so we can add it there. Also we want to do it after the points get their normal color (given by overlay, etc), so that in the case where we want the highlighted points to be a different shape instead of color (for example), we still have the opportunity to keep the points' original colors.

I ran into a bit of trouble with tooltips. Despite my best efforts, I couldn't hide the tooltips of the original data. I tried changing marker sizes, colors, lines, everything, but the tooltips still appeared. I finally found an option that I don't like because it removes some data, but at least the whole point isn't deleted from the trace.

Remaining To Dos:

• General scatterplot testing to make sure nothing broke
• Add prop that gives the highlight trace a specific name (would show up in the legend and hover text)

[asizemore]

Update: related backend PRs have been merged and are now deployed to QA.

[asizemore]

Ready for a look!
To test with highlights, use qa microbiome, DailyBaby scatterplot, and have at least one of the axes from the 16S entity. For example

After approval, I'll remove lines 163 and 2217 and test scatterplots again.

(This was hard for me to remember 😆 ) The goal is to have a solid processInputData function that does the highlighting so that our genomics components can use it, not for a fully functioning highlight scatterplot eda viz. So As long as we get highlights in the mbio scatter with our stubbed in point ids, we can be happy. Otherwise we'd be implmementing a lot of new logic in the eda scatterplot for a feature that won't be in the eda for some time.