Genes set the order that amino acids (the chemical building blocks of proteins) appear in the proteins which they code for. But, working from the gene, the form which the protein's 3-D structure will take cannot as yet be predicted. The extremely complex shapes in which the protein 'folds' has a profound effect on the properties it has within an organism.

The so-called 'protein folding problem' first described more than half a century ago, means that pharmaceutical and bio-informatic researchers (for example) are faced with very extensive problems when trying to design new medicines and enzymes - and also in understanding how currently known examples actually work.

We have little experimental knowledge of protein-folding energy landscapes.
[…]
We cannot consistently predict the structures of proteins to high accuracy. We do not have a quantitative microscopic understanding of the folding routes or transition states for arbitrary amino acid sequences. We cannot predict a protein’s propensity to aggregate, which is important for aging and folding diseases. We do not have algorithms that accurately give the binding affinities of drugs and small molecules to proteins. We do not understand why a cellular proteome does not precipitate, because of the high density inside a cell. We know little about how folding diseases happen, or how to intervene."

Source : The protein-folding problem, 50 years on in Science, vol. 338, 2012

Of the genes which are known to code for the generation of proteins, around 20% produce proteins which have functions that are as yet unknown. Oddly, this 20% figure seems to remain fairly constant for most of the organisms which have been studied. From the simplest yeast up to and including humans. See: The Royal Society Open Biology, 2019

Researchers using the DeepMind neural-network computing system have shown that it's possible to predict the folded protein shape with good accuracy - though it's currently not able to predict protein shapes from any arbitrary string of amino acids - the system relies instead on previous, experimentally-discovered, models of protein folds.

See : New Scientist , July 2022.

There is also a lack of understanding about how artificial neural networksplugin-autotooltip__plain plugin-autotooltip_bigNeural Networks

functionality_unexplained

""Multilayer neural networks are among the most powerful models in machine learning, yet the fundamental reasons for this success defy mathematical understanding."

Source : Proceedings of the National Academy of Sciences [open access]2018 115 (33)"

'Artificial Intelligence' (AI) systems predominantly use actually operate, and therefore how the results were achieved.

We have shown that the deep distance prediction neural network achieves high accuracy, but we would like to understand how the network arrives at its distance predictions and—in particular—to understand how the inputs to the model affect the final prediction.

Source : Nature Jan. 2020

Also see Protein Knottingplugin-autotooltip__plain plugin-autotooltip_bigProtein Knotting

Note: This article is an extension of the Protein Folding Problem

"Knotting in proteins was once considered exceedingly rare. However, systematic analyses of solved protein structures over the last two decades have demonstrated the existence of many deeply knotted proteins. and Intrinsically disordered proteinsplugin-autotooltip__plain plugin-autotooltip_bigIntrinsically disordered proteins

Before year 2000, it was generally assumed that the way in which proteins 'folded' was the sole key to understanding their function in life-systems. (See :Protein structuring ). Since then, it has been shown that many proteins do not entirely 'fold up' - leaving large sections of the protein chain as coils which appear to be random. This can profoundly affect the way in which they function and influence cellular systems.

Importance Rating