Assembly of a pan-genome from deep sequencing of 910 humans of African descent.
Sherman RM, Forman J, Antonescu V, Puiu D, Daya M, Rafaels N, Boorgula MP, Chavan S, Vergara C, Ortega VE, Levin AM, Eng C, Yazdanbakhsh M, Wilson JG, Marrugo J, Lange LA, Williams LK, Watson H, Ware LB, Olopade CO, Olopade O, Oliveira RR, Ober C, Nicolae DL, Meyers DA, Mayorga A, Knight-Madden J, Hartert T, Hansel NN, Foreman MG, Ford JG, Faruque MU, Dunston GM, Caraballo L, Burchard EG, Bleecker ER, Araujo MI, Herrera-Paz EF, Campbell M, Foster C, Taub MA, Beaty TH, Ruczinski I, Mathias RA, Barnes KC, Salzberg SL. · 2019
Read Assembly of a pan-genome from deep sequencing of 910 humans of African descent. by Sherman RM, Forman J, Antonescu V, Puiu D, Daya M, Rafaels N, Boorgula MP, Chavan S, Vergara C, Ortega VE, Levin AM, Eng C, Yazdanbakhsh M, Wilson JG, Marrugo J, Lange LA, Williams LK, Watson H, Ware LB, Olopade CO, Olopade O, Oliveira RR, Ober C, Nicolae DL, Meyers DA, Mayorga A, Knight-Madden J, Hartert T, Hansel NN, Foreman MG, Ford JG, Faruque MU, Dunston GM, Caraballo L, Burchard EG, Bleecker ER, Araujo MI, Herrera-Paz EF, Campbell M, Foster C, Taub MA, Beaty TH, Ruczinski I, Mathias RA, Barnes KC, Salzberg SL. (2019) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc6309586.txt.
Introduction
Sherman and co-authors build a pan-genome for people of African descent. Their study appeared in Nature Genetics. They deeply sequenced 910 individuals. Reads that failed to match the standard reference genome were kept. Those reads were assembled into new DNA stretches. The goal was DNA present in African populations but missing from the reference.
They found about 296 million base pairs of new sequence. It sat in 125,715 distinct stretches. That is roughly 10 percent more DNA than the reference holds. Some new stretches fall inside known protein-coding genes. Most lie between genes, with unknown roles. The work shows how one reference misses whole populations.
Catalog record
| CREATOR | Sherman RM, Forman J, Antonescu V, Puiu D, Daya M, Rafaels N, Boorgula MP, Chavan S, Vergara C, Ortega VE, Levin AM, Eng C, Yazdanbakhsh M, Wilson JG, Marrugo J, Lange LA, Williams LK, Watson H, Ware LB, Olopade CO, Olopade O, Oliveira RR, Ober C, Nicolae DL, Meyers DA, Mayorga A, Knight-Madden J, Hartert T, Hansel NN, Foreman MG, Ford JG, Faruque MU, Dunston GM, Caraballo L, Burchard EG, Bleecker ER, Araujo MI, Herrera-Paz EF, Campbell M, Foster C, Taub MA, Beaty TH, Ruczinski I, Mathias RA, Barnes KC, Salzberg SL. |
|---|---|
| YEAR | 2019 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 41 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Assembly of a pan-genome from deep sequencing of 910 humans of African descent By Rachel M. Sherman, Juliet Forman, Valentin Antonescu, Daniela Puiu, Michelle Daya, Nicholas Rafaels, Meher Preethi Boorgula, Sameer Chavan, Candelaria Vergara, Victor E. Ortega, Albert M. Levin, Celeste Eng Published in: Nature genetics DOI: 10.1038/s41588-018-0273-y ABSTRACT We used a deeply sequenced dataset of 910 individuals, all of African descent, to construct a set of DNA sequences present in these individuals but missing from the reference human genome. We aligned 1.19 trillion reads from the 910 individuals to the reference genome (GRCh38), collected all reads that failed to align, and assembled these reads into contiguous sequences (contigs). We then compared all contigs to one another to identify a set of unique sequences representing regions of the African pan-genome missing from the reference genome. Our analysis revealed 296,485,284 bp in 125,715 distinct contigs present in the African-descended populations, demonstrating that the African pan-genome contains ~10% more DNA than the current human reference genome. Although the functional significance of nearly all of this sequence is unknown, 387 of the novel contigs fall within 315 distinct protein-coding genes while the rest appear to be intergenic. Since its initial publication 1,2, the human genome sequence has undergone continual improvements aimed at filling gaps and correcting errors. The latest release, GRCh38, spans 3.1 gigabases (Gb) with just 875 remaining gaps 3. The ongoing effort to improve the human reference genome, led by the Genome Reference Consortium, has in recent years added alternate loci for genomic regions where variation cannot be captured by single nucleotide polymorphisms (SNPs) or small insertions and deletions (indels). These alternate loci, which comprise 261 scaffolds in GRCh38, capture a small amount of population variation and improve read mapping for some data sets. Despite these efforts, the current human reference genome derives primarily from a single individual 4, limiting its usefulness for genetic studies, especially among admixed populations such as those representing the African diaspora. In recent years, a growing number of researchers have emphasized the importance of capturing and representing sequence data from more diverse populations and incorporating these data into the reference genome and genomics studies in general 5–7. The alternate loci in GRCh38 offer one possible way to add such diversity, although it is unclear whether such a solution is sustainable as more populations are sequenced. Among other problems, the addition of alternate loci as separate contigs can mislead sequence alignment programs, which were designed under the assumptions that each read has a single true point of origin, and that the genome is represented as a linear haploid sequence 8. The lack of diversity in the reference genome poses many challenges when analyzing individuals whose genetic background does not match the reference. This problem may be addressed by using large databases of known SNPs (e.g., dbSNP 9), but this solution only addresses single-base differences and small indels, and is not adequate for larger variants. Findings from the 1000 Genomes Project indicate differences between populations are quite large; examining 26 populations across five continents, 86% of discovered variants were found to be present in only one continental group. In that study, the five African populations examined had the highest number of variant sites compared to the remaining 21 populations 10. One way to address the limitations of a single reference genome is to sequence and assemble reference genomes for other human sub-populations. The 1000 Genomes Project, Genome in a Bottle, and other projects have assembled draft genomes from various populations, including Chinese, Korean, and Ashkenazi individuals 11–15. Others have used highly homogenous populations (e.g., Danish, Dutch, or Icelandic individuals) together with assembly-based approaches to discover SNPs and structural variants (SVs), including up to several megabases of non-reference sequence common to these populations 16–19. While these variant analyses are a step in the right direction, to date none have produced a reference-quality genome that can replace GRCh38 3, although this is an explicit goal of the Danish Genome Project (see URLs). While efforts to produce new reference genomes are worthwhile, attempts to create a “pan-genome” of a human population; i.e., a collection of sequences representing all the DNA in that population, are rare. Although multiple pan-genomes have been created for bacterial species 20–22, as yet there are no pan-genomes for any other animal or plant species. The lack of pan-genomes is due in part to the technical challenges of assembling many deeply-sequenced genomes de novo and combining them into one genome. While the Danish Genome Project focused on 50 trios of non-admixed individuals (removing admixed samples from their study 17), our study focuses on a highly heterogeneous group of admixed individuals. Because the human reference genome is largely complete (i.e., the sequence has very few gaps), our strategy for creating a pan-genome focused on finding large insertions. This approach, although computationally demanding, made the African pan-genome assembly process described here feasible. A 2010 study that sequenced one Asian and one African individual used the novel sequences identified to estimate that a full human pan-genome would contain an additional 19–40 megabases (Mb) not in the current reference genome 11. Recent efforts to sequence a Dutch population and a set of 10,000 individuals have supported this estimate, reporting 4.3 and 3.3 Mb of non-reference sequences respectively 18,23, however neither study was designed with the primary goal of discovering long, non-reference sequences. A 2017 study, where two haploid human genomes (hydatidiform moles) were sequenced using long reads, estimated that a single diploid genome may differ by as much as 16 Mb from the reference genome 24. As we describe here, our analysis of 910 deeply sequenced individuals, all from the Consortium on Asthma among African-ancestry Populations in the Americas (CAAPA) 25, produced a much larger amount of novel sequence (i.e. sequence absent from GRCh38) in the African pan-genome spanning 296.5 Mb. Below we describe the methods used to identify and validate these sequences along with comparisons to other human sequences. The African pan-genome (APG) contigs have been deposited at NCBI as Accession PDBU01000000 to provide a better foundation for future analyses of African-ancestry individuals. In total, we discovered 296.5 Mb of novel DNA distributed across 125,715 sequences assembled from 910 individuals of African descent (Table 1, Supplementary Figure 1). We took steps to ensure contaminants and redundant contigs were removed, resulting in a non-redundant set of human contigs representative of the entire study group (Figure 1). After discovery, we called presence/absence for all African pan-genome (APG) sequences in each CAAPA sample. A total of 33,599 contigs with a combined length of 81,096,662 bases represented sequences present in at least two individuals in the CAAPA cohort. When alignments above 80% coverage and 90% identity to Chinese and Korean genome assemblies were also considered shared, the number of non-private insertions increased to 61,410, totaling 160,475,353 bases and leaving 64,305 singleton contigs, a ~51% singleton rate. Of the 125,715 APG sequences, 1,548 (total length 4.4 Mb) were anchored to a specific location in the primary GRCh38 assembly. On average each individual contained 859 of these inserted sequences, with a single sequence being shared among 6 individuals (Table 2). Placed contigs were shared among more individuals, 196 on average, as shared sequences were more likely to