I think Booker, Harris, Warren, Klobuchar, Bernie, Biden, Bloomberg, and Hillary are all among the most likely to run. Ojeda declared he was running yesterday but he voted for Trump in 2016, so... yeah. Beto might run too but I'd be surprised.
Generally incumbents have about 2:1 odds in generic contests as they have a lot of natural advantages. I think Trump has a much smaller margin for error. Any kind of economic slow down, even just a couple quarters of tepid performance, might do him in. That said even a 50.1% favorite is still a fav, and I think a lot on the left are discounting the fact that he can absolutely win again. He has some things going his way, too; wages grew faster than anytime in a decade this last 6 month period, and if high employment finally does put wage pressure out there, even with spiraling out of control inequality, it's probably enough to give him a small boost. His problem is his base is just not that big. There are more Dems than Republicans nationally and more Dems than Trump Republican voters nationally especially. He badly needs to either depress turnout of the former or find whatever magic pill Ohio has (their white rural vote expanded big time in these midterms and Republicans did really well in the state) and use that to keep the midwest. Michigan and Wisconsin and Pennsylvania will decide the election. Latest data wonks I saw had Michigan at 51-49 generic Dem vs. Trump when factoring in approval rating and demographics and stuff. Knowing this state could be the single most important state in 2020 is despair inducing.
Dems getting the senate is still a battle but I guess now that 2018 is going to go +1 or 2 as for Republicans as opposed to 3 or 4 like it originally looked, far from impossible. Trump's perpetual disapproval and potential massive turnout for 2020 probably means more Dem state gains.