Sequences become complete as copies and neighbours separate
Human and zebra finch genomes and a nanopore method show that new biological information can come from separating similar copies and neighbouring residues, not merely extending a read.
Science··Evening
The human genome becomes two separate texts
The Telomere-to-Telomere Consortium has announced a complete diploid reconstruction of HG002 that assembles the maternal and paternal copy of every chromosome separately. Work by teams at Johns Hopkins and NIST adds roughly 900 million DNA letters to the previous reference, amounting to 15 per cent more genome. The results appeared as a package of 12 papers in Cell and Cell Genomics. HG002 comes from one living donor and is used as reference material by measurement laboratories, so the assembly does not represent all human variation; it resolves the two chromosome copies of one person in detail. The same package also covers genomes from macaque, marmoset, zebra finch, rat, vole, horse, donkey, and giraffe. The report places the advance beside a large change in access: sequencing one human genome cost about 5 billion dollars in 2003 and now costs about 5,000 dollars. The added value lies in more than reading additional letters. Sequences inherited from two parents, which can be blended in a reference, are assigned to their separate positions. “The human genome” is consequently presented less as one consensus string and more as two complete copies carried together in an individual.[1]
Missing regions expand the songbird gene list
Erich D. Jarvis's laboratory at Rockefeller University assembled a telomere-to-telomere, fully phased diploid genome of the zebra finch and published the result in Cell. The new assembly adds 90 million previously missing base pairs, identifies 2,710 unknown genes, and resolves 11 tiny dot chromosomes. It also provides the female W chromosome in full for the first time. The zebra finch therefore becomes the second vocal-learning species after humans to receive a complete genome. The work reports that birds share an organised centromere architecture previously assumed to be specific to mammals. Because centromeres help chromosomes separate correctly during cell division, completing these regions does more than fill blank spaces; it changes the available comparison of chromosome organisation. The human and zebra finch results belong to the same Telomere-to-Telomere Consortium package. Together, they show that a genome's usefulness depends not only on total length but also on whether each copy, tiny chromosome, and difficult region is attached to the right place. The appearance of 2,710 genes after missing sequence was completed in the zebra finch also shows how the categories “unknown” and “outside the measurement” can overlap in genome research.[2], [1]
Separating neighbouring residues advances protein reading
EANPSeq, reported in Nature Communications, addresses a similar separation problem in protein sequences. The method digests peptides stepwise with an exopeptidase and continuously recognises the resulting fragments as they pass through a nanopore. With machine-learning assistance, it matches the fragment data against pre-built libraries of shortened and mutated peptides. The authors report that the approach can distinguish two adjacent identical residues and precisely locate post-translational modification sites within consecutive residues. They present the result as proof of concept and as a possible route towards single-molecule protein sequencing. Because EANPSeq does not read a sequence directly, its coverage remains dependent on what the peptide library contains, and the broader difficulty of reading proteins one residue at a time is not yet fully removed. Even so, the three developments meet at the same measurement lesson. Separating maternal and paternal copies in the human genome, recovering missing regions and tiny chromosomes in the zebra finch, and distinguishing identical neighbours in a protein all make a sequence that once looked like one line more informative. New biological information can come less from producing a longer string than from resolving which overlapping or similar component belongs at each position.[3], [1], [2]
Related columns
For more information on this topic, you can read the related columns.
