A running list of potential "Advanced Download Options"

### Context

There are a number of "intermediate" results that people might want to obtain from refine.bio (e.g., RNA-seq at the transcript-level, SCAN.UPC processed data prior to quantile normalization). We may want to offer these as part of "Advanced Download Options." On this ticket, we'll keep track of potential options as we think of them or as mentioned by users.

Related: #512

### Problem or idea

#### Summarize duplicate identifiers

Currently, we summarize ("squish") to the gene mean. We could offer other options such as median, max, or [something more complicated from the Troyanskaya lab](https://github.com/AlexsLemonade/refinebio/issues/330#issuecomment-400061952).

#### RNA-seq

* Questions around _individual experiments_ are tracked on #512
* Filtering -- i.e., removing genes with very low counts -- and how the desired behavior might change based on aggregation strategy

#### Normalization

* With or without quantile normalization
* [DBNorm](https://doi.org/10.1186/s12859-017-1912-5)?
* If cross-platform aggregation occurs, return only the highly expressed genes after quantile normalization.

#### Missing values and imputation

* Instead of retaining only genes present in every sample, choose what % of samples could be missing values before a gene gets dropped
* With or without KNNimpute

#### Batch correction

This one strikes me as particularly [potentially dangerous](http://dx.doi.org/10.1093%2Fbiostatistics%2Fkxv027), but we _could_ offer a selection of batch correction strategies: [SVA](https://doi.org/10.1371/journal.pgen.0030161), [ComBat](https://doi.org/10.1093/biostatistics/kxj037), [RUV family](https://doi.org/10.1093/biostatistics/kxr034) (See also [Parker, et al.](https://doi.org/10.1093/bioinformatics/btu375)).

### Solution or next step

We should start gathering some information about these options via chatting with folks at conference, possibly surveys, etc.

### New Issue Checklist

- [x] The title is short and descriptive.
- [x] You have explained the context that led you to write this issue.
- [x] You have reported a problem or idea.
- [x] You have proposed a solution or next step.

No implementation files, tests, or entry points are named. Start by reviewing related issues #512 and #330, then gather information through conference discussions or surveys about which advanced download options users need; done would be a documented, prioritized set of options with defined requirements.

Written by the indexing model from the issue text.