Yes. Follow this link.

https://www.r-bloggers.com/partial-least-squares-regression-in-r/

Thanks.

Have a look at these:

http://www.milanor.net/blog/performing-principal-components-regression-pcr-in-r/

https://www.r-bloggers.com/partial-least-squares-regression-in-r/

PLS can be done by R as earlier answers show, but to select variables may be better to use other methods like random forest etc. PLS is a good start though.

Thank you all for the suggestions. In fact I have tested various packages including pls for QSAR. The resource 'Applied Predictive Modeling' by Max Kuhn was indeed of great help. Hopefully I will comeup with the rest of the metrices required to prove the predictive abilities of the models.