This is an interesting question. To answer it, we will need to think about a few things. Let's first think about what a combustion reaction is. A combustion reaction has the following general form:

Fuel source + O₂ --> (oxidized by-products) + heat

The fuel source is often something kind of organic molecule (something with C, H, and O), but metals and other gases can also combust. The oxidized product in organic molecule combustion is usually CO₂, and for metals they might look like some metal oxide compound.

So what? Well, let's think about two aspects of the reaction: the thermodynamic aspect, and the kinetic one. If we look at a combustion reaction thermodynamically, the products usually have much lower free energy than the reactants, so we have a highly exothermic reaction (i.e. lots of heat is released, which often helps to perpetuate the combustion as long as there is still fuel and oxygen remaining). Now if we think about it, lots of things could react to form products with much lower free energy. Why don't they? That's because there is a kinetic aspect to the problem as well. The kinetic aspect tells us how fast (and hence how likely) a reaction is to occur. The kinetics of a reaction are related to how large of an activation energy barrier the reaction has. The larger the activation energy, the more energy is required to "start" the reaction, and that energy usually comes in the form of thermal energy ("heat") or light.

How does thermal energy facilitate a combustion reaction? As you increase the temperature, the molecules in the system "jiggle around" more -- in the case of gases, you would see more translations, rotations, bond wiggling within a molecule, as well as collisions between molecules. Because of all this extra motion, the molecules can access more orientations/configurations, and are more likely to get into the appropriate orientation for the chemical reaction to occur. What does that mean? In any chemical reaction, such as a combustion reaction, where bonds are being broken and formed between different atoms in molecules, the parts of the molecules that undergo bond breakage and formation need to line up with each other. That's because there needs to be the appropriate overlap in the orbitals that are "donating" and "accepting" electrons.

So back to your original question: if there are enough molecular collisions occurring because conditions are right (temperature is high enough), or some other source of energy for activation is available, then combustion can begin to occur. And if the conditions remain favorable (enough thermal energy, sources of fuel and oxygen), then combustion may persist. However, it is not simply the fact that the molecules are colliding enough that combustion initiates. The molecules must be colliding in the appropriate orientations, and because they are colliding more frequently, they are more likely able to "access" those appropriate configurations. I hope this helps!

Chemical reactions are sometimes modeled by "Collision Theory," which is an important part of what you're asking!

Collision theory says that in order for a chemical reaction to happen, three separate things need to happen:
1. The molecules have to hit each other
2. The molecules have to hit each other in the right way (both have to be facing the right way)
3. The molecules have to hit each other with enough speed (energy of motion, or "kinetic energy") to activate the reaction

If all three things happen, then a chemical reaction occurs! From what I understand, you are asking if two molecules can hit each other hard enough (with enough energy) to catch on fire. That's an interesting question, and one that requires a quick look at what "fire" is. Fire is the chemical reaction that happens when oxygen, heat, and fuel combine. Fire needs all three in order to happen. I will describe a scenario where your question applies: Imagine a tank of propane (the gas, not liquid) and oxygen mixed together. At normal temperatures, there will not be fire because there isn't enough heat energy (it's still a very dangerous scenario!) However, if you heat up the container so much that the molecules of oxygen and propane hit each other very very hard, then the gasses will ignite, and you will have your fire!

So, the short answer to your question: yes! The long answer: fire needs the right conditions (oxygen, fuel, heat) and how fast or hard molecules hit each other is related to heat (that is, heat energy becomes kinetic energy in gasses). If you heat oxygen and fuel up enough, the molecules will hit each other with enough energy to self-ignite.

So technically, a flame is caused by molecules "striking" each other in the case of combustion of a material with oxygen. However, this is because fire is a chemical reaction and thus requires the collision of molecules. Though it doesn't make sense to think of this reaction like two metal balls hitting each other making a spark because sparks are made of colliding gas molecules themselves. Most chemical reactions don't result in a flame, but rather different molecular species than those that entered the reaction. Additionally, most collisions between gas molecules don't result in reaction. A lot of times to try and understand gases, we think of them like hard spheres like billiard balls that bounce around striking each other. However, this really is incorrect because what causes a colliding molecule to rebound when it hits another molecule is the electrical repulsion as the molecules get too close. Therefore, a "collision" between gas molecules is really more like when you try and push together magnets with the same pole facing each other and they push each other away.