Nothing that we know of, that is why it is called 'dark'.
I think the short answer is no, within current accepted theory. The most concise representation of General Relativity is,
Gμν = - κ Tμν
where Tμν is the energy-momentum tensor and by implication since mass is energy, it accounts for mass as well. Gμν is the Einstein curvature term and κ is Einstein's constant of gravity which is directly related to Newton's constant G. So curvature of space is directly linked the amount of mass and energy in a locality. Because of the relation E = mc^2, mass is by far the largest contributor to the curvature of space.
However, the point you make at the start of your post, 'We observe that the visible matter of galaxies is moving in ways that cannot be attributed to only the gravitational forces associated with this visible matter', is not backed up by all observational data!
Are you sure? Light going into a black hole increases the gravity of the hole and therefore "space time warping" [link is not peer reviewed, arXiv.org]. Photons of light do not have mass.
This is not useful information for dark matter. Dark matter is clearly not made of photons.
Everyone is going to laugh about my opinion in that if we could calculate the frame drag of the galaxy using earths frame drag as a representation and apply that figure with regards to the amount of visible matter contained within our galaxy would we get a figure that is close to what we have calculated for the amount of dark matter needed? Maybe I’m in left field on this or not I don’t know.
Frame dragging is just too tiny an effect compared to what would be needed.
This is not actually true.
In fact, all straightforward applications of the dark matter hypothesis, which looked promising at first, have pretty much been ruled out by observational evidence. The plain vanilla model in which there is a single type of thermal dark matter with a mass O(1-100) GeV has been ruled out by observation for almost a decade. And, there are, in fact, modified gravity theories (although not the most well known example of the genre called MOND) that do fit the data better than any of the extant dark matter theories (see, e.g., Moffat's MOG theory), and do so with fewer free parameters in their models, although few modified gravity theories have been tested as rigorously and by as large a group of investigators as the leading dark matter theories have been.
This is not to say that an entirely satisfactory and well tested solution on any front exists. But, a lot of the data points which folk wisdom assumes destroyed modified gravity theories (e.g. the Bullet Cluster) do no such thing. Indeed, data points like the Bullet Cluster actually do more harm to dark matter particle theories than to modified gravity theories (many of which can accommodate this observation).
Also, just to be clear, there is really no reasonable doubt that phenomena usually attributed to dark matter, that can not be explained with GR (at least as currently interpreted and applied*) and ordinary matter, exist and are pervasive. The phenomena attributed to dark matter can only be explained with some sort of new physics that either involves beyond the Standard Model particles, or forces that have effects different from GR as currently interpreted and applied, or both. These phenomena are by far the most compelling direct observational evidence that the "Core Theory" of GR plus the Standard Model is not complete and that New Physics are necessary to explain what is observed. (In contrast, "dark energy" phenomena can be completely explained to the limits of experimental observation with the cosmological constant of conventional GR.)
* There are a couple of promising gravitation based theories that claim that they do not actually modify GR but involve a means of applying GR-like concepts different than the way that the vast majority of researchers in the field apply GR to the analysis of complex systems operationally.
I'm still not convinced that MACHOs are ruled out.
Just recently we observe a compact object from deep space flying past the solar system.
Why not calculate the frame drag associated with the black hole (or holes) in the center of a galaxy based on its mass and revolution speed (earths frame drag has been observed). Then determine both the value and relative position of the wake created by the frame drag/revolutions and plug that into a computer model to see if that can fill the void that is suggesting dark matter is needed. I believe some kind of estimate can be derived using the dynamics associated with a so called perfect liquid (or what I would call cold plasma)?
Doing so for the black hole of a similar mass to the one at the center our galaxy and figuring the framing dragging effect at, say, 50,000 ly from the center, it works out to being the equivalent of an additional 7.4e-54 km/sec. Besides, frame dragging falls off with distance from the mass, so any effect it would have would be stronger near the BH than it is further, But stellar speeds nearer the center of the galaxies aren't the problem, it's the ones on the outskirts.
Does anyone claim MACHOs do not exist? Someone could claim MACHOs are 80% of the Milky Ways mass, or 1%, or 0.001%. If it is 0.001% as comets/asteroids that would be 109 solar mass. Something like 1026 comets. Finding one of them will not prove much.
1/'Oumuamua was not a halo object.
Allow me to remind you of some of the pertinent evidence which rules out MACHOs.
As noted by stefan r above, there is no doubt that some MACHO candidates exist (although primordial black holes have not yet been observed and there is good reason to doubt that they exist), but there simply are enough of them and they aren't in the right places to account for a meaningful share of dark matter phenomena. The smallest primordial black holes are impossible:
The constraints on non-primordial black hole (PBH) MACHOs (terrestrial planets, baby or ordinary gas giants, neutron stars, cannibalized white dwarfs that become helium or diamond planets) are severe and these candidates have been ruled out for many years. Indeed, massive compact halo objects (MACHOs) are pretty much ruled out, in general, as dark matter candidates (citations in the original omitted and paragraph breaks added in the quotation below):
There have been arguments that PBHs are not subject to some of the exclusions above applicable to other kinds of MACHOs. But, a the MACHO exclusion chart below (which is a few years out of date (with the left side showing the ratio of MACHOs to dark matter and the horizontal axis showing MACHO mass) is as follows:
In terms of MACHOs are a dark matter candidate, only the very top part matters, since 10-2 means just 1% of dark matter can be accounted for by MACHOs of that type. You really need to be above 10-1 (i.e. 10%) to be considered as a significant source of dark matter phenomena.
As this chart shows, there are pretty significant observational constraints on the size of primordial black hole dark matter, however (relying mostly on this source). The sweet spot is 10^22 kilograms, which is a bit less than the mass of the Moon (which is 7*10^22 kilograms), plus or minus, which would imply a typical primordial black hole with an event horizon radius of about 0.1 millimeters.
Another paper on PBHs was even less bullish:
Fabio Capela, et. al, "Constraints on primordial black holes as dark matter candidates from capture by neutron stars." Phys. Rev. D 87, 123524 (2013) (link is to open access pre-print version conformed to final print version).
Since the chart above was made the observational constraints on PBHs have further tightened. For example:
Daniele Gaggero, et al., "Searching for Primordial Black Holes in the radio and X-ray sky" (Pre-Print December 1, 2016).
Another recent paper constraining MACHOs as cluster dark matter (keeping in mind that galaxy clusters are inferred to have much more dark matter proportionately than any kind of galaxy) finds:
Theodorus Maria Nieuwenhuizen "Subjecting dark matter candidates to the cluster test" (October 3, 2017) (omitting from the abstract conclusions about sterile neutrino dark matter candidates).
It also turns out that the model used in many PBH exclusion analyses is actually too lenient and that more realistic modeling makes the exclusions even tighter.
Separate names with a comma.