For science as a whole, 2017 was the best of times and the worst of times.

The year saw the approval of three gene therapies, a clinical milestone that has its basis in modern molecular biology. Earlier-stage research and clinical trials, meanwhile, saw progress in other gene therapies and gene editing that could be applicable to a broad swath of genetic diseases.

But the support for using scientific principles to understand the world, and scientific evidence to inform policy, is being eroded from the highest political office in the land.

In addition to its Breakthrough of the Year – which featured gene therapy as one of the runners-up – Science magazine publishes a Breakdown of the Year annually. This year's top choice was the "epic estrangement" between President Trump and the scientific community.

As far as that estrangement is concerned, the outlook for biomedical science is less bleak than for some other areas. The Science piece noted that "Trump's decision to retain National Institutes of Health (NIH) Director Francis Collins and NSF Director France Cordova has given many scientists hope that academic research will remain relatively unscathed during his administration." And other Trump choices – Scott Gottlieb as head of the FDA, Brenda Fitzgerald as director of the CDC, and Jerome Adams as U.S. surgeon general – "are generally viewed as mainstream appointees who understand and support their agency's work."

The fact that it now bears mentioning when an appointee understands and supports their agency's work is a stark illustration of how dismal the overall position of science is in the current administration.

The most recent illustration that a competent appointee is no guarantee of freedom to operate was last week's dust-up over the supposed banning of seven words – diversity, entitlement, evidence-based, fetus, science-based, transgender and vulnerable – in the CDC's budget request to Congress.

CDC Director Fitzgerald denied that there are "banned" words in the budget request. And indeed, according to Science, each word appears in the FY 2018 budget request at least once, and three of them – diversity, fetus and science-based – appear more frequently than they did in the three previous years.

There is a sharp decrease, however, in the use of the other four phrases. Most troubling is the decreased use of the phrase evidence-based, which could be found 125 times in 2017's budget request and only 38 times in 2018's, possibly indicating self-censorship.

The administration's hostility to both the phrase and the concept of evidence-based is likely to be self-reinforcing.

Science quoted Arden Bement, who was director of the National Institute of Standards and Technology and the NSF under President George W. Bush, as saying that "unfortunately, providing science advice to a president who resists advice, would not understand it, or would distort it for personal and political reasons would be futile and frustrating."

And now for the good news

For those who cared to look, 2017 also brought both triumphs for the evidence-based approach and progress in gathering evidence.

The FDA approved its first gene therapy in 2017. And its second, and its third: Kymriah (tisagenlecleucel, Novartis AG), Yescarta (axicabtagene ciloleucel, Gilead Sciences Inc.) and Luxturna (voretigene neparvovec-rzyl, Spark Therapeutics Inc.).

In response to Kymriah's approval, Emily Whitehead, who was the first person to be treated with the CAR T-cell therapy in 2012, was named one of Nature's "Ten People Who Mattered in 2017."

Science mentioned the approvals, but focused on the earlier-stage gene therapy AVXS-101 (Avexis Inc.) for spinal muscular atrophy type 1. In November, the company published data from a pivotal phase I trial in The New England Journal of Medicine. All 15 patients treated with AVXS-101 were alive and had reached at least 20 months of age at the data cutoff, and none of them were in need of ventilation. And of 12 patients in the high-dose cohort, 11 were able to control their head movements, sit at least briefly, and speak. Two children learned to walk independently. Historical controls have a 20-month survival rate of less than 10 percent.

Earlier in December, Avexis announced that "on Dec. 5, 2017, the company had an end-of-phase I meeting with FDA with respect to AVXS-101 for SMA type 1. The company anticipates providing an update on feedback from FDA following the receipt of the final meeting minutes in early January."

Both Science and Nature also highlighted the 2017 development of so-called base editors as an important advance in genome editing.

In October, researchers from the Broad Institute reported they had developed a method of converting the base adenosine (A), which is one of the four base pairs in DNA, to an inosine, which is read by the transcription machinery as a guanine (G). In 2016, the same team had already identified a method for converting CG to TA base pairs.

As yet, not all base changes can be corrected via editing, and for that and other reasons, the method is a good ways away from any clinical approach. But base conversion could ultimately provide a strategy for a very large number of small genetic changes; 32,000 of the more than 50,000 human genetic changes that are associated with disease risk are point mutations, where the high-risk associated variant differs from the low-risk associated variant by just one base pair.