What is the standard way to work with a diploid reference genome? Complementary strands?

Question

At the moment, the standard reference genomes (e.g. hg19, hg38) are haploid genomes. We know that the human genome is diploid. Naturally, the latter would be the respectively correct representation of the human genome.

More and more biologists are using emerging technologies into order to capture the diploid nature of genetic information, e.g. phasing SNPs between mother and father chromosomes.

What is the standard way that bioinformaticians generated a standard diploid reference genome?

Actually, reference genomes are not truly haploid (by my understanding). Given that reference genomes are 5'-3' coordinated, in order to create a complementary strand, one would need to take the 3'-5' complement. In order to have a diploid genome, you need two reference genomes and two 3'-5' complements.

More importantly, how have large-scale genomic studies dealt with the fact that the haploid reference genome is a consensus-based "half" of a human genome?

Answer 1

For calling small variants, the standard way is to simply call diploid genotypes. You can already do a variety of research with unphased genotypes. You may further phase genotypes with imputation, pedigree or with long reads/linked reads, but not many are doing this because phasing is more difficult, may add cost and may not always give you new insight into your data. For these analyses, we use a haploid genome. For human samples, the vast majority of "large-scale genomic studies" are done this way.

A diploid reference actually doesn't help much with reference-based analysis; it only complicates algorithms. What could help a lot is a population reference, which may be represented by a graph or a compressed full-text index or both. In theory, if you have a comprehensive population reference and a capable mapping algorithm, you may call extra variants that would not be callable with short reads. In practice, however, there are quite a few technical challenges. Handling population references is a research topic. There are no "standards" yet.

If the goal is to assemble a new reference genome from a diploid sample, we almost always prefer to produce a diploid assembly. Unfortunately, I believe there are no "standard" procedures, either. SuperNova from 10x genomics builds the diploid information into a graph. Falcon from PacBio uses "unzip". I don't think they have got widely used and evaluated so far.

PS: saw your edit while writing the above. The fact that the genome only represents one strand does not mean we have to create the complement strand explicitly in analyses. We do most of reverse complement on the fly in algorithms as well as in mind.

reference genomes are not truly haploid

That depends on how the reference is assembled. If you sequence a haploid sample (e.g. bacteria), your assembly will be haploid. If you sequence an inbreed lab strain that is almost homozygous (e.g. mouse and fruit fly), your assembly will be nearly haploid. If you sequence a diploid sample, your assembly is very likely to be a mosaic of the two haplotypes. In case of the human reference genome, it is more complicated. It is largely a mosaic of several humans by stitching ~150kb haplotypes from these samples.

Answer 2

At the moment, the standard reference genomes (e.g. hg19, hg38) are haploid genomes. We know that the human genome is diploid. Naturally, the latter would be the respectively correct representation of the human genome.

The premise of the OP's question is false. The natural reference representation of the human genome is not diploid.

Think of a reference genome as a map, and not as a specific example of a human being's DNA.

Not only is the human genome reference haploid, but it is also a composite genome. This means that the human genome reference sequence is composed of sequences from multiple individuals. In other words, the human reference does not correspond to any one human sequence.

Any particular read from a DNA sequencer will be a read from a human genome that diverges from the reference genome. So an algorithm that tries to match the read to the reference genome will always need to handle potential discrepancies. Adding a second map against which to match a read would not change that fact. Therefore, there is little value in providing a second haploid reference genome.

Side note: There are parts of the human genome that "are too complex to be represented by a single path", and the Genome Reference Consortium provides "alternate loci" for such regions of the genome.

Answer to the OP's second question

More importantly, how have large-scale genomic studies dealt with the fact that the haploid reference genome is a consensus-based "half" of a human genome?

I interpret this question as "How do large genomic studies represent genome diversity that cannot be represented by a haploid reference genome?"

One standard approach that has been very popular in genome-wide association studies is the use of a haplotype reference panel such as for example from the 1000 Genomes Project. Modern phasing programs exploit the shared ancestry of the samples in a study and reference haplotypes to phase a study's samples. One popular phasing program is Impute2

Answer 3

To add to all the other great answers, I would mention that the question is somewhat misleading. If the reference genome is for a single individual, then it should be diploid. However, it's a reference for all humans. It should really contain billions of copies to fully account for all the diversity. Since that is not realistic, the reference serves as a simple approximation.

This point was addressed by the recent Korean genome paper:

Human genomes are routinely compared against a universal reference. However, this strategy could miss population-specific and personal genomic variations, which may be detected more efficiently using an ethnically relevant or personal reference. ... Systematic comparison of human assemblies shows the importance of assembly quality, suggesting the necessity of new technologies to comprehensively map ethnic and personal genomic structure variations.

Answer 4

There are some assemblers that produce assembly graphs that attempt to describe all the possible haploid paths within a set of reads. Such an assembly attempts to capture all the diploid variation (and/or population variation) in a sample at the expense of not having full-length chromosomes.

Canu (for example) will produce contigs that are extended as long as consensus is maintained across different reads, but when there is a reliable break in coverage (i.e. an area where chromosomes are heterozygous) then the contigs will be broken up. Canu provides as output a GFA file (assembly graph) that can be used to determine which paths might combine together into a single chromosome.

Answer 5

Since nobody has tackled you second question, I'll give it a go:

The complementary strand to a genome sequence is the 5'-3' reverse complement, not the 3'-5' complement. This sequence is not captured in standard reference genomes, but the information is. Given a sequence, it's trivial to calculate it's reverse complement, and any tool that's designed to work with sequence data takes this feature into account.