They are two different sides of the same coin.  Here's a drawing of the data lifecycle in meteor:

![Meteor Data Lifecycle][1]

* [**Publish**][2] - Which data is sent from the *server*  
* [**Subscribe**][3] -  Which data the *client* requests publications for
* [**Methods**][4] - How to manipulate data from the client on the server
 * **Note** - this will typically be run on both on the client and the server.  The client will make a prediction as to what the server will do so it can update right away.  Then *latency compensation* will kick in when the method is run on the server and the canonical decision is made.




[1]: http://i.stack.imgur.com/qoaCi.png
[2]: http://docs.meteor.com/#/full/meteor_publish
[3]: http://docs.meteor.com/#/full/meteor_subscribe
[4]: http://docs.meteor.com/#/full/meteor_methods

What is the difference between **PubSub** and **Methods** in Meteor?!

Can I put Methods in Server folder like Publishs?

To me seen like the same, but Methods is more reactive.
