There is a problem in using language to predict relations among people.  
Language is a culturally transmitted and not a biological trait.  Just because  
a native of Hong Kong speaks English does not mean that he or she is  
necessarily of British decent.  In fact it is more likely that he or she is Chinese. 

On the other hand, DNA is the genetic material that determines biological inheritance.  "Lum examined DNA that is found within mitochondria (mtDNA),  
small cellular bodies that function as the energy factories and storehouses of our cells. Mitochondria are inherited from the body of the mother's fertilized egg, and are transmitted maternally to the next generation" (Lum & Heathcote, 1998, 1).  This analysis ignores inheritance from a father. 

Lum wanted to find out if the DNA similarities would agree with the linguistic similarities. "For instance, would there be close mtDNA similarity among allthe groups who speak Oceanic Austronesian (OCAN) languages?  In Micronesia, OCAN languages include those spoken in the Southwest Islands of Palau,  
Yap Outer Islands, Chuuk, Pohnpei, Kosrae, Nauru, the Marshalls, and  
Kiribati.  As well, all Polynesian languages are classified as OCAN, as are 
many languages spoken within Melanesia.  In contrast, people from Yap  
Proper, Palau (other than the Southwest Islands) and the Marianas speak  
languages that are classified as non-Oceanic Austronesian (AN)" Lum &  
Heathcote, 1998, 2).  Would Chamorus, Palauans and Yapese mtDNA be  
closer to AN speakers from Southeast Asia or closer to OCAN-sepaking Micronesians (Lum & Heathcote, 1998, 2)? 

Lum analyzed the hair of 455 people.  He estimated the genetic distances  
among the various population groups by comparing the mtDNA sequences 
of each group.  He compared the mtDNA of people from Guam, Rota, Saipan,  
Palau, the Southwest Islands of Palau, Yap Proper, and Kapingamarangi with  
those from "479 volunteers from other parts of Micronesia, Polynesia, Melanesia, Australia, Indonesia, the Philippines, mainland Southeast Asia, and East  
Asia" (Lum & Heathcote, 1998, 2). 

There was a general mtDNA agreement with linguistic classifications among  
all the OCAN speakers.  But the Western Micronesian AN speakers' (Yapese 
and Palauans) mtDNA is most similar to OCAN-speaking Micronesians.  The Chamorus were very different.  These AN speakers clustered "with a diverse grouping of AN speakers and even with speakers of non-Austronesian 
languages" (Lum & Heathcote, 1998, 5) 

"The Chamorro sample has closest mtDNA similarity to two aboriginal Malay  
groups from western coastal Thailand (Moken and Urak Lawoi).  These results  
are intriguing, as all three groups are non-Oceanic Austronesian (AN) speakers.  Surprisingly, the next closest degree of similarity to the Marianas sample is with Japan, then aboriginal Australians, then a sample from Java" (Lum & Heathcote, 1998, 5). 

What does all this mean?  "First, the generally close agreement between  
mtDNA and linguistic relationships indicates that common historical  
processes were involved in the dispersal of both maternal lineages of people, 
and languages that they speak, throughout the Pacific.  Fourteen of fifteen  
Polynesian and Micronesian samples (mostly OCAN-speakers) cluster  
together and have relatively close mtDNA ties with samples from the Philippines, Borneo and South China (Canton)"  (Lum & Heathcote, 1998, 5). 

This shows that Micronesians and Polynesians have a southeast Asian 
homeland.  "In contrast, studies based on DNA contributed by both females  
and males to their offspring generally indicate a greater degree of Melanesian heritage for Polynesians and Micronesians. 

The results for Palau and Yap are not so tidy.  The "mtDNA and linguistic relationships do not agree?  Lack of such agreement can mean a number of  
things, but mixing of populations - after initial linguistic settlement - springs 
to mind first.  Western Micronesia and Melanesia are regions where a greater amount of such mixing is indicated" (Lum & Heathcote, 1998, 5-6).  This  
"suggests that Palau has been 'seeded' by people with ancestral roots in island Southeast Asia and Melanesia, as well as the more easterly parts of Micronesia" (Lum & Heathcote, 1998, 6). 

Chamoru mtDNA is very distinctive when compared to other Micronesians and 
Polynesians.  This suggests that the Marianas have a different settlement history than the rest of Micronesia. Chamorus have not mixed much with other Micronesians. This does not mean that Chamorus are Malays.  "What suchclose mtDNA affinity suggests is that Chamorros and aboriginal Malays hav 
common maternal ancestors, 'way back when'.  The 'way back when' time being before the Chamorros were a distinctively crystallized group, before the  
colonization of the Marianas by people whose descendants would only 
later develop the way of living that defined  them as 'Chamorros'" 
(Lum & Heathcote, 1998, 7). 

References 

Bellwood, P. S.  "The Colonization of the Pacific:  Some Current Hypotheses."  
In The Colonization of the Pacific:  A Genetic Trail. Edited by Adrian V. S. Hill and Susan W. Serjeantson. Oxford: Clarendon Press, 1989. 

Lum, J. Koji and Gary M. Heathcote. "Genetic Relationships of Micronesian Populations: A Project Update." Anthropology Resource & Research Center Non-Technical Report Series, no. 1. Mangilao, Guam: University of Guam, 1998. 

Murdock, George P. "Genetic Classification of the Austronesian Languages:  
A Key to Oceanic Culture History." In Peoples and Cultures of the Pacific, ed. Andrew  P. Vayda. Garden City New York:  The Natural History Press, 1968. 

Olmo, Richard K.  Personal Communication, 1998. 

Russell, Scott.  Personal Communication, 1997. 

Topping, Donald, Pedro M. Ogo, and Bernadita C. Dungca.  Chamorro- English Dictionary. 
Pacific and Asian Linguistic Institute Language 
Texts:  Micronesia.  Honolulu:  University Press of  Hawaii, 1975.